51 of 51 criteria have a sourced finding; 42 were also cross-checked by a second agent. Every comparison that includes Claude Code links here for scope, limitations, atomic answers and sources.
Humans approve the rubric, templates and prototype; they do not review every finding. An agent source cross-check is not a hands-on product test. A missing sourced finding is not evidence that a capability is absent.
Claude Code has CLI, Desktop, IDE, browser and mobile clients. Execution is local or cloud by surface; mobile is a client only. Desktop Linux is beta; 3P Desktop has separately documented provider and Console-key routes.
Scope
Surfaces: Claude Code CLI, Claude Code desktop, Claude Code web/mobile, Claude Code IDE extensions · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Pro, Max, Ultra, Team and Enterprise. Desktop, web and mobile need a claude.ai subscription, except the Desktop exception: gateway routing can be configured in the app or by an administrator; Enterprise deployments can route Desktop to Google Cloud Agent Platform or a gateway via managed settings; and Claude Desktop on 3P runs the Code tab on Amazon Bedrock, Google Cloud Agent Platform, Microsoft Foundry or a self-hosted LLM gateway (feature-availability). · OS: macOS, Windows, Linux · Execution modes: local, managed-cloud, remote-control · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
Not separately cross-checked
Capabilities and authentication differ across CLI, Desktop, IDE, mobile and cloud.
No installed-client or authenticated scenario was run.
Desktop 3P supports the documented Console-key route; mobile/cloud still require eligible claude.ai accounts.
Mobile apps [supported] Claude iOS and Android apps (iPad uses the iOS app) have a Code tab for cloud sessions, projects, Remote Control and Dispatch; code never runs on the phone. Sources: claude-code:mobile-nov07.
Mobile browser [supported] A phone browser can open the Remote Control session at claude.ai/code and send instructions. An eligible subscription and an awake local host are required. Sources: claude-code:gap-mobile-web.
Cli [supported] CLI: macOS 13+, Windows 10 1809+/Server 2019+, Ubuntu 20.04+, Debian 10+, Alpine 3.19+, x64 or ARM64, 4 GB+ RAM; the built-in sandbox works on macOS, Linux and WSL2 only. Sources: claude-code:setup-nov07, claude-code:sandboxing-nov07.
Editor extensions [supported] VS Code extension (VS Code 1.94.0+, also Cursor) and a JetBrains plugin that drives the separately installed CLI. Sources: claude-code:vscode-nov07, claude-code:jetbrains-nov07.
Task actions [partial] Mobile reaches cloud sessions, Remote Control local sessions, or Desktop Dispatch. Dispatch is on Pro/Max/Ultra; the host must stay awake for local execution. Sources: claude-code:mobile-nov07, claude-code:platforms-nov07.
Notifications [supported] Push notifications to the phone when Remote Control is active (task finished or decision needed), Dispatch notifications, terminal bell/desktop notification settings in the CLI, and a Notification hook. Sources: claude-code:mobile-nov07, claude-code:settings-nov07, claude-code:hooks-nov07.
Cross device continuity [partial] Cloud sessions persist across devices; Remote Control continues a local CLI/Desktop/VS Code session from phone or browser while the machine stays on; /teleport pulls a cloud session into the terminal; /desktop continues a CLI session in Desktop. A terminal session cannot be pushed to the cloud (handoff is one-way); a desktop-to-phone move was not tested. Sources: claude-code:mobile-nov07, claude-code:remote-control-nov07, claude-code:web-nov07, claude-code:commands-nov07.
Claude Code embeds its own harness and runs Claude models only. Providers: Anthropic (subscription or Console), Amazon Bedrock, Google Cloud Agent Platform, Microsoft Foundry, Claude Platform on AWS, and gateways that expose an Anthropic-format API. Anthropic does not support routing to non-Claude models and the docs state that workers in every multi-agent mode are Claude sessions.
Scope
Surfaces: Claude Code CLI · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Claude Code
Availability
generally-available
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Scope is the Claude Code harness; unofficial third-party routing to other model families is not covered by the documentation and was not tested.
5 atomic answers
Harnesses [supported] One harness: Claude Code itself. Subagents, agent view, teams, workflows and projects all run Claude sessions. Sources: claude-code:agents-nov07.
Custom adapters [partial] Gateways that expose a supported Anthropic API format are supported; other tools can be added as MCP servers. No adapter mechanism for running a different agent harness inside Claude Code was found in the pages checked. Sources: claude-code:llm-gateway-nov07, claude-code:agents-nov07.
Session integration [supported] Claude Code is the harness; it can be driven programmatically through claude -p, the Python/TypeScript Agent SDK, or as a stdio MCP server (claude mcp serve). Sources: claude-code:headless-nov07, claude-code:sdk-nov07, claude-code:mcp-nov07.
Mixed harness operation [unsupported] Docs state "In every approach the workers are Claude sessions. To involve a different tool, expose it to Claude as an MCP server", so mixed-harness operation is not offered natively. Sources: claude-code:agents-nov07.
Pro, Max, Ultra, Team and Enterprise include Claude Code. Console/API and third-party cloud-provider routes are separate. Free excludes Code; platform API credits bundled with Max/Ultra cannot pay Code usage.
Scope
Surfaces: Claude Code CLI, Claude Code desktop, Claude Code web/mobile, Claude Code IDE extensions · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Pro, Max, Ultra, Team, Enterprise, Console API, third-party providers · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
generally-available
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Local-model statement is the vendor support position; unofficial base-URL workarounds were not tested.
No trial or promotional pricing beyond the public pricing page was checked.
Two Console sign-ins without API keys are not isolated by config directory (authentication page); the account_switching atom is partial for that case.
Api accounts [supported] Anthropic Console API key or Console login, plus Bedrock, Google Cloud Agent Platform, Microsoft Foundry and Claude Platform on AWS credentials. Sources: claude-code:auth-nov07, claude-code:feature-availability-nov07.
Local models [unsupported] Anthropic states it does not support routing Claude Code to non-Claude models through any gateway; only Claude models are documented. Sources: claude-code:llm-gateway-nov07.
Account switching [partial] Separate CLAUDE_CONFIG_DIR directories keep settings, session history and claude.ai logins separate, so several accounts can stay signed in; /login and /logout switch accounts. Two Console sign-ins without API keys are not isolated by config directory because their credentials are stored outside it (authentication page). Sources: claude-code:auth-nov07, claude-code:support-pro-max-nov07.
Vendor account requirement [supported] A claude.ai subscription, Console account or third-party provider credential is required. Desktop, cloud sessions and mobile need a claude.ai subscription, except the Desktop exception: gateway routing, Enterprise managed-settings routing to Agent Platform or a gateway, and the 3P Desktop Code tab on Bedrock, Agent Platform, Foundry or a self-hosted gateway (feature-availability). Sources: claude-code:overview-nov07, claude-code:feature-availability-nov07.
Provider restrictions [supported] Cloud sessions, Remote Control, teleport, Slack need claude.ai login and are not reachable with an API key or Bedrock, Agent Platform or Foundry. Desktop needs claude.ai login except the partial exception: gateway routing, Enterprise managed-settings routing, and the 3P Desktop Code tab on Bedrock, Agent Platform, Foundry or a self-hosted gateway. Channels in research preview permit claude.ai or Console API-key authentication, with organization controls; third-party providers remain excluded. Sources: claude-code:feature-availability-nov07, claude-code:remote-control-nov07, claude-code:channels-nov07.
CLI onboarding is one install command plus browser login, and works in any directory. Worktree isolation needs git; cloud sessions need GitHub (or a local bundle). Self-hosted runners exist for Team/Enterprise. Remote Control is off by default on Team and Enterprise plans until an Owner enables it.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: Current first-party documentation read 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local, managed-cloud, self-hosted-runner · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
Not separately cross-checked
Onboarding was read from documentation, not timed hands-on.
Self-hosted environments are public beta and exclude ZDR and HIPAA organizations.
7 atomic answers
First task setup [supported] Install with one command (curl/irm/brew/winget), run claude in a project and log in; the first task can start immediately. Sources: claude-code:overview, claude-code:auth.
No git [partial] The local CLI runs in any directory, but --worktree needs a git repository (hooks can substitute) and parallel background sessions in a non-git directory edit the same files without isolation. Sources: claude-code:worktrees, claude-code:agent-view.
Local git without remote [supported] For cloud sessions, claude --cloud bundles a local repository with no remote (bundle under 100 MB); local CLI use needs no remote. Sources: claude-code:web.
Hosted git [partial] Cloud sessions clone from GitHub (GitHub App or /web-setup), GitHub Enterprise Server on Team/Enterprise; GitLab and Bitbucket repos only as local bundle; GitLab merge request URLs work with --worktree and claude agents, and GitLab CI/CD is documented. Sources: claude-code:web, claude-code:ghe, claude-code:whats-new, claude-code:overview.
Prerequisite guidance [supported] System requirements and install troubleshooting are documented; Git for Windows is recommended on native Windows. Sources: claude-code:setup, claude-code:overview.
Managed cloud setup [supported] Browser onboarding connects GitHub. Pro/Max/Ultra get a Default environment automatically; Team/Enterprise can need the environment-creation form unless an Owner enabled Quick setup. Sources: claude-code:web-quickstart, claude-code:web, claude-code:cloud-env-nov07.
Byo vps onboarding [partial] Self-hosted environments (public beta, Team/Enterprise) run cloud sessions on customer runners; Remote Control server mode on an own always-on machine is documented for Pro and above. Sources: claude-code:selfhost, claude-code:remote-control.
Programmable access is strong on the CLI: claude -p with JSON output, Python/TypeScript Agent SDK, claude mcp serve and HTTP/webhook triggers for routines. The routine /fire endpoint is experimental and several surfaces are research preview.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
API/webhook coverage is limited to routines and the SDK; no hands-on integration test.
Api [partial] Routines expose an HTTP /fire endpoint with per-routine bearer tokens under an experimental beta header and for claude.ai users only; a general REST API for session control was not found in the pages checked. Sources: claude-code:routines-nov07.
Authentication [supported] Subscription login, ANTHROPIC_API_KEY or provider credentials for the CLI/SDK; bearer token per routine API trigger. Sources: claude-code:auth-nov07, claude-code:routines-nov07.
Claude Code is an MCP client (http, sse, stdio, websocket) with OAuth and scoped config, supports plugins/marketplaces, skills, hooks, custom subagents, reads AGENTS.md, and can import Codex, Gemini CLI and Cursor configuration.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
generally-available
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Integration ownership (which connectors Anthropic maintains) was not established.
Direct AGENTS.md reading requires Claude Code v2.1.277 or later, and in some sessions Claude cannot read AGENTS.md; the docs suggest importing it from CLAUDE.md there.
v2.1.295 negotiates MCP protocol revision 2026-07-28 by default; MCP_PROTOCOL_NEGOTIATION=legacy opts out (changelog, 2026-10-08). The stored answers for this claim do not address protocol versions.
7 atomic answers
Mcp client [supported] MCP client for http, sse, stdio and websocket servers, plus claude.ai connectors when logged in with a subscription. Sources: claude-code:mcp.
Plugins [supported] Plugins bundling skills, agents, hooks and MCP servers installed from marketplaces with /plugin. Sources: claude-code:plugins.
Instructions import [supported] By default, reads AGENTS.md directly when neither CLAUDE.md nor CLAUDE.local.md exists in the working directory or above, from v2.1.277. /import brings Codex, Gemini CLI and Cursor configuration, but is unavailable on Bedrock, Agent Platform, Foundry, Claude Platform on AWS, Claude apps gateways or when feature-flag fetching is disabled. Sources: claude-code:memory, claude-code:commands.
Authentication [supported] OAuth 2.0 for remote MCP servers via /mcp or claude mcp login; header and environment-variable credentials for others. Sources: claude-code:mcp, claude-code:cli.
Permission scope [supported] Local, project and user scopes; project servers need approval; managed MCP allow/deny lists; per-tool permission rules. Sources: claude-code:mcp, claude-code:admin.
Integration owner [unknown] Integration owner: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
The mobile app is a client for cloud or Remote Control sessions, and IDE extensions wrap or bundle the CLI engine. Desktop framework and inspected-source evidence were not established.
Scope
Surfaces: Claude Code CLI, Claude Code desktop, Claude Code web/mobile, Claude Code IDE extensions · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: Not specified · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Desktop UI framework and CLI implementation language were not established from first-party sources; the public repository holds issues, plugins and the changelog under an all-rights-reserved license, and no source inspection was done.
4 atomic answers
Desktop implementation [unknown] Desktop implementation: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Mobile implementation [partial] Architecture only: the mobile app is a client for cloud and Remote Control sessions and no code runs on the phone. Its UI framework (native, React Native or web) was not established. Sources: claude-code:mobile.
Hybrids [partial] Documented architecture: the VS Code extension bundles a copy of the CLI (its version equals the bundled Claude Code version), the JetBrains plugin runs a separately installed CLI, and the Desktop app reads the same settings files as the CLI. Framework hybrids (for example Electron with native shells) were not established. Sources: claude-code:vscode, claude-code:jetbrains, claude-code:desktop, claude-code:platforms.
Implementation evidence [unknown] Implementation evidence: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
No native issue tracker was found in the documentation, changelog or repository checked; work arrives by prompt, Slack/Claude Tag mentions, GitHub @claude mentions and GitHub-event routines, and external trackers such as Linear or Notion are reached through MCP connectors. Authoritative-tracker, ID and sync semantics depend on the connector and were not established.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: Current first-party documentation read 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
official-integration
Checked
2026-11-07 (15 days old)
Agent cross-check
Not separately cross-checked
A native issue/backlog tracker was not established; session Task tools are a separate documented checklist; the pages checked do not describe tracker identity or sync.
9 atomic answers
Local creation [supported] When Task tools are available (default on Opus 4.7, Sonnet 4.6, Haiku 4.5 and earlier, opt-in on other models, always in background and cloud sessions), TaskCreate adds product-owned tasks to the session task list; setting CLAUDE_CODE_TASK_LIST_ID stores and shares the list in a named directory under ~/.claude/tasks/. Sources: claude-code:gap07-tasks, claude-code:gap07-task-persistence.
External import [partial] Linear, Notion and other trackers are reachable by adding MCP servers or claude.ai connectors; this is user-configured, not a built-in importer. Sources: claude-code:platforms, claude-code:mcp.
Chat git triggers [supported] Claude in Slack (earlier version) answers @mentions only from Pro, Max and Ultra accounts in workspaces not connected to Claude Tag. Team and Enterprise use Claude Tag. The legacy bot was retired effective 2026-10-05 for workspaces an organization connected to Claude Tag. GitHub Actions @claude mentions on issues and PRs, and routine triggers on GitHub events. Sources: claude-code:platforms, claude-code:slack-second-review-nov07, claude-code:ghactions, claude-code:routines.
Authoritative tracker [unknown] Authoritative tracker: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Stable ids [partial] TaskGet retrieves a specific task; named task-list directories can be reused across sessions. Cross-system issue identity was not established. Sources: claude-code:gap07-tasks, claude-code:gap07-task-persistence.
Sync [unknown] Sync: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Supported edits [unknown] Supported edits: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
External roundtrips [unknown] External roundtrips: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Plan mode proposes a plan before edits and plans can be saved into the repository; /goal tracks a completion condition. Durable plan objects with revisions and acceptance-criteria editors were not found.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
Not separately cross-checked
Plan files are written to ~/.claude/plans or a project-relative plansDirectory; plan revision history, amendment tracking and links from plans to implementation were not established.
7 atomic answers
Plans [partial] Plan permission mode explores read-only and proposes a plan for approval; plan mode writes plan files to ~/.claude/plans or a project-relative plansDirectory, and the docs suggest saving the plan to the repo and committing it for later execution. Sources: claude-code:permission-modes, claude-code:web, claude-code:cc-plansdir-chk1.
Acceptance criteria [partial] /goal sets a verifiable completion condition that a model re-checks after each turn. Sources: claude-code:goal.
Editors [supported] The plan approval dialog offers Ctrl+G to edit the proposed plan in the default text editor. Sources: claude-code:gap07-plan-editor.
Revisions [unknown] Revisions: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Implementation links [unknown] Implementation links: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Amendment [unknown] Amendment: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Restart persistence [partial] Sessions are saved as local transcripts and restored with --continue/--resume; session-scoped scheduled tasks are restored on resume unless expired; in-process agent-team teammates are not restored. Sources: claude-code:sessions, claude-code:scheduled, claude-code:teams.
Ordinary TaskUpdate changes task dependencies. Task tools are available by default on selected older model families, can be enabled on other models, and are provided on every model in background and cloud sessions. Experimental agent teams add a shared task list where unresolved dependencies block claiming. Hierarchy and epics were not established.
Scope
Surfaces: Claude Code CLI · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Agent teams are experimental and off by default; no epic/story hierarchy, task-to-epic movement or identity retention was found.
9 atomic answers
Epics [unknown] Epics: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Hierarchy [unknown] Hierarchy: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Blockers [partial] Agent-team tasks can depend on other tasks; a pending task with unresolved dependencies cannot be claimed. Sources: claude-code:teams-nov07.
Readiness [partial] Teammates self-claim the next unassigned, unblocked task. Sources: claude-code:teams-nov07.
Enforcement [partial] Experimental team members do not self-claim a dependency-blocked task, but the user or lead can still assign one. This is not a universal dispatch block. Sources: claude-code:teams-nov07.
State propagation [partial] Task status can lag because teammates sometimes fail to mark tasks completed, which blocks dependents (documented limitation). Sources: claude-code:teams-nov07.
Task to epic [unknown] Task to epic: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Identity retention [unknown] Identity retention: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Workflow constraints [unknown] Workflow constraints: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Persistent instructions and memory are documented: CLAUDE.md at managed/user/project/local scope, AGENTS.md reading, skills, imports and rules, and editable per-project auto memory stored locally.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
generally-available
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Scope verified from documentation only; no hands-on takeover test.
Direct AGENTS.md reading requires Claude Code v2.1.277 or later, and in some sessions Claude cannot read AGENTS.md; the docs suggest importing it from CLAUDE.md there.
10 atomic answers
Instructions [supported] CLAUDE.md files at managed, user, project and local scope with imports and .claude/rules. By default, AGENTS.md is read when neither CLAUDE.md nor CLAUDE.local.md exists in the working directory or above. Sources: claude-code:memory.
Skills [supported] Skills (SKILL.md, Agent Skills standard) load on demand and can be shared via plugins. Sources: claude-code:skills, claude-code:plugins.
Memory [supported] Auto memory saves learnings to markdown under ~/.claude/projects/<project>/memory/. Sources: claude-code:memory.
Scope [supported] Memory is per project (shared across worktrees of one repo); instructions have managed, user, project and local scopes. Sources: claude-code:memory.
Provenance [partial] The interface notes "Saved N memories"/"Recalled N memories" and /memory lists loaded files; per-fact source attribution was not found. Sources: claude-code:memory.
Correction [supported] Memory files are plain markdown that can be edited at any time, including via /memory. Sources: claude-code:memory.
Deletion [supported] Auto memory files can be deleted at any time; auto memory can be disabled globally or per project. Sources: claude-code:memory.
Freshness [partial] Memory files are excluded from the transcript retention sweep; no staleness detection was found. Sources: claude-code:memory.
Takeover continuity [partial] Project CLAUDE.md can be committed and shared; auto memory lives on the individual machine under ~/.claude, so a colleague does not receive it unless files are shared. Sources: claude-code:memory.
Code understanding relies on agentic search (grep, glob, file reads) plus optional LSP code-intelligence plugins and user-supplied MCP indexes; images can be attached; web search and fetch tools exist. A built-in semantic index and citation or conflict-detection features were not found.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
generally-available
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
No retrieval-quality or citation test was run.
10 atomic answers
Code search [supported] Claude searches with grep/glob and file reads; code-intelligence plugins add language-server navigation. Sources: claude-code:large-codebases.
Indexing [partial] Documentation points to LSP plugins and exposing an organization search or RAG index as an MCP tool; no built-in persistent index was described. Sources: claude-code:large-codebases.
External knowledge [partial] WebFetch and WebSearch tools plus MCP servers provide external knowledge. Sources: claude-code:mcp, claude-code:platforms.
Freshness [partial] Claude reads files fresh on each tool call, so edits made elsewhere are seen the next time it reads. Sources: claude-code:common-workflows.
Api retrieval [unknown] Api retrieval: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Citations [unknown] Citations: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Conflict detection [unknown] Conflict detection: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Review enforcement [unknown] Review enforcement: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Claude Code has persistent native task checklists with optional cross-session lists and experimental team assignment/dependencies. A full issue backlog is not established.
Scope
Surfaces: Claude Code CLI · Version: Task tools documented 2026-11-07; model-dependent or opt-in since v2.1.268; team assignment experimental · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
generally-available
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Task tools are absent by default on newer interactive models unless enabled; this is a checklist rather than a full issue tracker. Team assignment requires experimental teams.
17 atomic answers
Create [partial] TaskCreate adds a native checklist task when Task tools are enabled; this is not a full issue backlog. Sources: claude-code:gap07-tasks.
Edit [partial] TaskUpdate changes task details and status, dependencies or deletes a task. Sources: claude-code:gap07-tasks.
Discuss [unknown] Discuss: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Prioritize [unknown] Prioritize: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Label [unknown] Label: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Assign [partial] Experimental team leads can assign tasks and teammates can self-claim unblocked tasks. Sources: claude-code:gap07-team-tasks.
Search [unknown] Search: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Filter [unknown] Filter: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Archive [unknown] Archive: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Reopen [unknown] Reopen: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Project epic views [unknown] Project epic views: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Idea parking [unknown] Idea parking: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Explicit dispatch [unknown] Explicit dispatch: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Task session separation [supported] Tasks live in ~/.claude/tasks/<list>/ separately from session transcripts and background process lists; named lists can be shared across sessions. Sources: claude-code:gap07-task-persistence, claude-code:gap07-team-tasks.
Durable states [supported] Team tasks have pending, in-progress and completed states, and task-list directories remain on disk after sessions end. Sources: claude-code:gap07-team-tasks.
External tracker requirement [partial] Native checklist tracking uses product Task tools and disk storage rather than an external tracker. Complete issue/backlog management remains unestablished. Sources: claude-code:gap07-task-persistence, claude-code:gap07-tasks.
TaskCreate, TaskGet, TaskList and TaskUpdate manage session checklists. They are available by default on selected older model families, can be enabled on other models, and are provided on every model in background and cloud sessions. Experimental agent teams add shared claiming and assignment. A native shared issue tracker was not established.
Scope
Surfaces: Claude Code CLI · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Agent teams are experimental and disabled by default. Local foreground sessions provide task tools by default only on selected model families; other models can opt in. Background and cloud sessions provide them on every model. These are session task lists, not a native issue tracker.
Update [supported] TaskUpdate changes status, dependencies and details or deletes tasks; availability depends on model/client as described in the summary. Sources: claude-code:tools-reference-nov07.
Claim [supported] Teammates self-claim unblocked tasks with file locking to avoid races. Sources: claude-code:teams-nov07.
Assign [supported] The lead can assign tasks explicitly to teammates. Sources: claude-code:teams-nov07.
State changes [supported] Tasks move pending, in progress, completed. Sources: claude-code:teams-nov07.
Children [unknown] Children: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Results [unknown] Results: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Human visibility [partial] Humans can view the task list with Ctrl+T and message teammates. Team configuration is removed when the session ends, but its task list is preserved for resumed sessions. Sources: claude-code:interactive-nov07, claude-code:teams-nov07.
Attribution [unknown] Attribution: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Permissions [partial] Task tools depend on model, opt-in and tool permissions; team members lacking them coordinate with messages instead. Task-specific ACLs are not established. Sources: claude-code:gap07-tasks, claude-code:gap07-team-tasks.
Conflicts [partial] Team task claiming uses file locks against simultaneous claim races. Conflicting task edits were not tested. Sources: claude-code:gap07-team-tasks.
Living product specifications were not found as a native feature; instructions files and plans in the repository are the nearest documented mechanisms.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
These are general instruction files, not a specification object with approvals or drift detection.
9 atomic answers
Persistent requirements [partial] CLAUDE.md and committed plan files persist instructions in the repository. Sources: claude-code:memory, claude-code:web.
Human edits [supported] Humans edit CLAUDE.md and memory files directly. Sources: claude-code:memory.
Agent edits [partial] Claude can write to memory and project files, subject to permissions. Sources: claude-code:memory.
Revisions [unknown] Revisions: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Approvals [unknown] Approvals: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Implementation links [unknown] Implementation links: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Later retrieval [unknown] Later retrieval: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Drift detection [unknown] Drift detection: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Review integration [unknown] Review integration: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Parallel work is supported locally (multiple terminals, worktrees, subagents, agent view background sessions, Remote Control server mode with default capacity 32) and in the cloud; every parallel session draws on the same subscription limits.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Agent view is research preview; no concurrency stress test was run, so observed_concurrency is unknown.
5 atomic answers
Concurrent sessions [supported] Several sessions can run at once in separate terminals or worktrees, as background sessions in agent view, or as cloud sessions; Remote Control server mode defaults to 32 concurrent sessions. Sources: claude-code:agents, claude-code:agent-view, claude-code:remote-control.
Quotas [partial] Parallel sessions share one set of subscription limits; the docs say ten agents use quota roughly ten times as fast; agent view is local and stops if the machine shuts down. Sources: claude-code:agent-view, claude-code:web.
Observed concurrency [unknown] Observed concurrency: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Status visibility [supported] claude agents shows each session as working, needing input or completed; /tasks lists background work in a session; /workflows lists workflow runs. Sources: claude-code:agent-view, claude-code:agents.
Agent teams (experimental) document lead assignment and file-locked task claiming; outside teams, assignment is the user dispatching sessions.
Scope
Surfaces: Claude Code CLI · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Agent teams are experimental and disabled by default; one team per session.
Routing [unknown] Routing: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Duplicate prevention [partial] File-locked task claims prevent two teammates from claiming one task. Agent teams do not isolate teammate files: two teammates editing the same file leads to overwrites, so the docs say to give each teammate its own files. Worktrees isolate separate sessions. Sources: claude-code:teams-nov07, claude-code:agents-nov07.
Ownership enforcement [partial] Experimental team task claims use file locking to prevent simultaneous claims of the same task. File write ownership is not established. Sources: claude-code:gap07-team-tasks.
Delegation is inspectable through subagents (own model, effort, tools), agent teams, cross-session messaging and dynamic workflows. All workers are Claude sessions; agent teams are experimental and off by default.
Scope
Surfaces: Claude Code CLI · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Claude Code
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Agent teams are experimental; in-process teammates are not restored on resume.
Authentication [unknown] Authentication: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Live progress [partial] Agent view rows, /tasks and /workflows show run state. A background subagent result arrives later as a completion notification, and Claude reports the subagent as still running if asked about progress first. Sources: claude-code:agent-view-nov07, claude-code:agents-nov07, claude-code:subagents-nov07.
Retained results [unknown] Retained results: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Invocation [supported] Claude spawns subagents with the Agent tool, users @-mention named agents, or start a session with --agent. Sources: claude-code:subagents-nov07.
Cloud scheduled tasks, formerly routines, support schedule/API/GitHub triggers independently of the laptop. Desktop local schedules and session /loop have separate uptime and expiry rules.
Scope
Surfaces: Claude Code CLI, Claude Code desktop, Claude Code web/mobile, Claude Code IDE extensions · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Scheduled tasks: Pro, Max, Ultra, Team, Enterprise · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Routines are research preview with experimental API; no scheduled run was executed.
Completion conditions [partial] /goal keeps a session working until a model judges the condition met; routines are prompt runs with no documented completion condition. Sources: claude-code:goal-nov07, claude-code:routines-nov07.
Overlaps [unknown] Overlaps: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Retries [unknown] Retries: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
In the CLI a human can interrupt, stop, queue steering messages and resume sessions; background sessions show Needs input state; push notifications and hooks surface pending decisions. Some commands differ on mobile.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local, remote-control · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Pause semantics beyond interrupt/stop, escalation policy and error visibility were not tested.
10 atomic answers
Pause [unknown] Pause: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Pending decisions [supported] Agent view groups sessions under Needs input and the peek panel shows the pending question. Sources: claude-code:agent-view.
Escalation [unknown] Escalation: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Visible errors [unknown] Visible errors: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Chat terminal parity [partial] Remote Control sessions in the app support most but not all commands (for example /plugin and /resume are local-only). Sources: claude-code:mobile, claude-code:remote-control.
Retained history [supported] CLI transcripts are retained locally (30 days by default). Transcripts of sessions started or most recently continued in Claude Desktop or Cowork are exempt from the 30-day cleanup by default (data-usage page). Cloud session history is restored on reopen. Sources: claude-code:sessions, claude-code:data-usage, claude-code:web, claude-code:data-usage.
Agent view (research preview) is a terminal fleet screen that groups background sessions under Needs input, Ready for review, Working and Completed, with filters, peek and attach; Desktop has a session sidebar. Dependency graphs and parent-child epic views were not found.
Scope
Surfaces: Claude Code CLI · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
beta
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Agent view is research preview and shows only local background sessions; no fixture with many agents was run.
8 atomic answers
Fleet fixture [unknown] Fleet fixture: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Parent child [unknown] Parent child: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Dependencies [unknown] Dependencies: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Attention [supported] Sessions needing input are grouped and counted at the top of agent view; the prompt footer counts agents waiting. Sources: claude-code:agent-view-nov07.
History [supported] Completed sessions stay listed; claude agents --all includes completed background sessions. Sources: claude-code:agent-view-nov07.
Workload [unknown] Agent view rows show each session state and one-line report. A done/total count was replaced by one-line reports in v2.1.205, and no per-epic workload view was found. Sources: claude-code:agent-view-nov07.
Navigation errors [unknown] Navigation errors: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Fixture mode [unknown] Fixture mode: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Quick tasks are one prompt away; larger efforts use subagents, agent teams, projects (cloud beta) or dynamic workflows, configured in natural language or markdown. Agent teams are experimental.
Scope
Surfaces: Claude Code CLI · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Orchestration surfaces are spread across several preview features; no end-to-end scenario run.
Growing epic [partial] Projects (public beta on Pro, Max and Ultra, rolling out gradually) provide one long-running coordinating conversation that starts threads; threads are marked done by you, by Claude, or automatically after a week with no activity. Agent teams are one team per session. Sources: claude-code:projects-nov07, claude-code:teams-nov07.
Agent configuration [supported] Custom subagents are markdown files with tools, model, effort and isolation. Sources: claude-code:subagents-nov07.
In run changes [partial] Humans can message teammates and stopped subagents can be resumed; changing a running workflow script mid-run was not described. Sources: claude-code:teams-nov07, claude-code:subagents-nov07.
Live roles [unknown] Live roles: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Retained relationships [unknown] Retained relationships: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Discoverability [partial] /workflows, /tasks and claude agents list orchestration state in separate places. Sources: claude-code:agents-nov07.
Auth friction [unknown] Auth friction: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Claude Code runs locally by default; cloud sessions run on Anthropic VMs (4 vCPU, 16 GB, 30 GB) with no separate compute charge; cloud sessions are available on Pro, Max, Ultra and Team plans, and for Enterprise users with premium seats or Chat + Claude Code seats, optionally on customer-run self-hosted runners (Team/Enterprise beta); Remote Control lets other devices drive a local session.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Cloud: Pro, Max, Ultra, Team, Enterprise; self-hosted: Team, Enterprise (public beta) · OS: Not specified · Execution modes: local, managed-cloud, self-hosted-runner, remote-control · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Region and exact idle-expiry durations were not found.
Self hosted [supported] Self-hosted environments run cloud sessions on customer runners (public beta, Team/Enterprise, not with ZDR/HIPAA). Sources: claude-code:selfhost-nov07.
Hybrid [partial] Projects threads can run on the user computer via Remote Control; claude --cloud and /teleport move work between terminal and cloud (one direction each). Sources: claude-code:projects-nov07, claude-code:web-nov07.
Provisioning [supported] Fresh VM per session with a cached setup-script snapshot. Sources: claude-code:cloud-env-nov07.
Payer [supported] Cloud VM use has no separate compute charge; it consumes the account subscription limits. Sources: claude-code:web-nov07.
Region [unknown] Region: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Expiry [supported] Idle VMs pause after a few minutes and can later be reclaimed; reopening provisions a fresh VM with history restored. Sources: claude-code:cloud-env-nov07, claude-code:web-nov07.
Other device continuation [supported] Cloud sessions are reachable from browser and phone; Remote Control exposes a local session to claude.ai/code and the app. Sources: claude-code:mobile-nov07, claude-code:remote-control-nov07.
Isolation options: git worktrees per session, an OS-level sandbox for shell commands (macOS/Linux/WSL2, off by default), dev containers or the sandbox runtime for the whole process, and a fresh VM per cloud session. The CLI by default runs with the user permissions and file tools outside the sandbox.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: Current first-party documentation read 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: macOS, Linux, WSL2 · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
generally-available
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
No escape or boundary test was run.
12 atomic answers
Checkout [supported] --worktree creates a separate checkout and branch. Sources: claude-code:worktrees.
Branch [supported] Worktrees use a new branch named worktree-<name> by default. Sources: claude-code:worktrees.
Os sandbox [supported] Built-in sandbox on macOS, Linux and WSL2; native Windows runs commands unsandboxed. Sources: claude-code:sandboxing.
Filesystem [supported] Sandbox writes are limited by default to the working directory, a per-user temp directory and directories added with --add-dir; reads of most of the machine, including credential files, are allowed unless denied. Sources: claude-code:sandboxing.
Process [partial] Sandbox applies to shell commands and their child processes; file tools, MCP servers and hooks run outside it. Sources: claude-code:sandboxing.
Network [supported] Sandboxed commands reach hosts only through a domain-allowlist proxy; cloud environments have network access levels. Sources: claude-code:sandboxing, claude-code:cloud-env-nov07.
Secrets [partial] Sandbox inherits environment secrets by default; credential protection settings exist; cloud network secrets are available on Pro, Max, Ultra, Team and Enterprise, subject to the documented role and configuration requirements. Sources: claude-code:sandboxing, claude-code:cloud-env-nov07.
Code separation [partial] Self-hosted runners lock to a single session owner, running only that owner’s sessions until the runner exits. Cloud sessions separately receive fresh VMs; no boundary test ran. Sources: claude-code:selfhost-nov07, claude-code:cloud-env-nov07.
Security boundary [partial] Sandbox is off by default and Anthropic recommends containers, VMs or the sandbox runtime for stronger isolation. Sources: claude-code:sandboxing, claude-code:sandbox-env.
Sessions persist as local transcripts and resume; cloud sessions restore history after VM reclamation but lose background processes; Remote Control reconnects after sleep, and server mode gives up after roughly 10 minutes without network; background local sessions survive sleep but not shutdown.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local, managed-cloud, remote-control · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
No crash or restart experiments were run; claims are from documentation.
12 atomic answers
Client disconnect [supported] Cloud sessions keep running. Remote Control reconnects after a brief network loss; in server mode Claude Code gives up after roughly 10 minutes and the claude remote-control process exits. Local sessions need the machine. Sources: claude-code:web, claude-code:remote-control.
Worker crash [unknown] Worker crash: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Service restart [unknown] Service restart: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Expiry [supported] Cloud VMs are reclaimed after inactivity; conversation history is restored on reopen but running background work is not. Sources: claude-code:web.
Transcript reload [supported] Local transcripts reload with --continue/--resume; /rewind restores code and conversation checkpoints. Sources: claude-code:sessions, claude-code:checkpointing.
State reconstruction [partial] Background subagents, shell commands and workflows that ended with the previous process appear as notes in the resumed transcript; in-process teammates are not restored. Sources: claude-code:sessions, claude-code:teams.
Context loss [partial] Compaction can lose detail; a 2.1.293 fix addressed Claude redoing finished work after compaction. Sources: claude-code:changelog.
Duplicate actions [unknown] Duplicate actions: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Manual repair [unknown] Manual repair: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Multi day continuity [partial] Local CLI transcripts are kept 30 days by default (cleanupPeriodDays); memory files are excluded from the sweep. Transcripts of sessions started or most recently continued in Claude Desktop or Cowork are exempt from the 30-day cleanup by default (data-usage page). Sources: claude-code:data-usage, claude-code:memory, claude-code:data-usage.
Cloud environments offer setup scripts, a filesystem-snapshot cache, network access levels and pre-installed toolchains; local runs inherit the host. Process state, ports and automatic repair are not covered.
Scope
Surfaces: Claude Code web/mobile · Version: Current first-party documentation read 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Ubuntu 24.04 (cloud VM) · Execution modes: managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Cloud-only findings; the local CLI inherits whatever the host provides.
10 atomic answers
Dependencies [supported] Setup scripts install dependencies before Claude Code launches; SessionStart hooks cover project setup everywhere. Sources: claude-code:cloud-env-nov07, claude-code:hooks.
Caches [supported] Environment cache keeps installed packages and pulled images when the setup script finishes in about five minutes; rebuilt on changes or after about seven days. Sources: claude-code:cloud-env-nov07.
Snapshots [supported] Anthropic snapshots the filesystem after setup and reuses it for later sessions. Sources: claude-code:cloud-env-nov07.
Services [partial] docker compose images can be pre-pulled, but cache stores files only, not running processes. Sources: claude-code:cloud-env-nov07.
Ports [unknown] Ports: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Secrets [partial] Environment variables are visible to anyone using the environment; network secrets are available on Pro, Max, Ultra, Team and Enterprise, subject to role and configuration requirements. Sources: claude-code:cloud-env-nov07.
Duplicate installs [unknown] Duplicate installs: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Collisions [partial] Worktrees isolate code between parallel sessions; CLAUDE_JOB_DIR gives each background session its own temporary directory for scratch files. Sources: claude-code:worktrees, claude-code:agent-view.
Remote setup [supported] The cloud environment dialog takes name, network level, variables and setup script. Sources: claude-code:cloud-env-nov07.
Automatic repair [unknown] Automatic repair: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Limits include print-mode --max-turns and --max-budget-usd (client-side estimate), subagent maxTurns, tool timeouts, plan usage limits with automatic continue, organization spend limits, 10 automatic retries and --fallback-model.
Scope
Surfaces: Claude Code CLI · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
generally-available
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
Not separately cross-checked
--max-budget-usd and --max-turns apply to print mode only; subagent maxTurns caps individual subagents in any session.
8 atomic answers
Time limits [unknown] Time limits: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Spend limits [supported] --max-budget-usd in print mode, usage-credit spend limits for Pro/Max, and organization spend limits for Team/Enterprise. Sources: claude-code:cli-nov07, claude-code:costs-nov07.
Timeouts [supported] Bash defaults to a two-minute foreground timeout, up to ten minutes. On timeout it moves work to the background except for sleep-prefixed commands, which stop. Unattended background work has a separate documented timeout; interactive background work differs. MCP calls can abort after an idle window. Sources: claude-code:cloud-env-nov07, claude-code:mcp-nov07.
Retries [supported] Transient API failures are retried up to 10 times with exponential backoff. Sources: claude-code:errors-nov07.
Operational statistics are available via /usage, /insights, OpenTelemetry metrics, logs and traces (beta), and the Team/Enterprise analytics dashboard; failure reasons are documented in the error reference.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Failure status [unknown] Failure status: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Approval status [unknown] Approval status: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Reconnect consistency [unknown] Reconnect consistency: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Performance method [unknown] Performance method: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Responsiveness [unknown] Responsiveness: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Cost views: /usage session cost (client-side estimate at list price unless an administrator's managed modelPricing table sets contracted rates), plan usage bars, OTel cost and token metrics, Console and Team/Enterprise dashboards; no separate compute charge for cloud VMs. Estimates can differ from the bill.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Pro, Max, Team, Enterprise, Console API · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Costs reconcile against provider billing only outside Claude Code.
14 atomic answers
Task cost [partial] Per-session cost resets on /clear; per-task cost is not a separate object. Sources: claude-code:costs-nov07.
Session cost [supported] /usage Session block shows total cost computed locally from tokens. Sources: claude-code:costs-nov07.
Child cost [partial] Usage is broken down by model and attributed to subagents and skills in the plan usage breakdown. Sources: claude-code:costs-nov07.
Epic cost [unknown] Epic cost: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Harness aggregation [unknown] Harness aggregation: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Billed vs estimated [supported] The figure is locally computed from token counts at list price unless an administrator sets managed modelPricing rates. It remains an estimate; Console usage is authoritative for billing. Sources: claude-code:costs-nov07.
Tokens [supported] Input, output and cache tokens by model are shown. Sources: claude-code:costs-nov07.
Subscription allowance [supported] Plan usage bars and activity stats show subscription limits. Sources: claude-code:costs-nov07.
Provider reconciliation [partial] Console usage page and spend reports are the billing source. Sources: claude-code:costs-nov07.
Compute cost [supported] Cloud VMs carry no separate compute charge. Sources: claude-code:web-nov07.
Refresh delay [unknown] Refresh delay: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Compute-aware scheduling is limited: Remote Control and self-hosted runners cap concurrency, Bash commands have timeouts and can be cancelled, and subscription limits are shared. Workflow fan-outs run up to 16 agents by default and fewer on CPU-limited hosts, configurable from 1 to 256 (v2.1.269 or later). RAM-aware queues, check deduplication and fairness were not found.
Scope
Surfaces: Claude Code CLI · Version: Current first-party documentation read 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local, self-hosted-runner · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
Not separately cross-checked
Workflow concurrency adapts to available CPUs; RAM-aware scheduling was not established.
12 atomic answers
Concurrency [partial] Workflow fan-outs run up to 16 agents by default, fewer when Claude Code has fewer CPUs, including inside a CPU-limited container; CLAUDE_CODE_WORKFLOW_MAX_CONCURRENT_AGENTS sets 1 to 256 (v2.1.269 or later) (workflows). Remote Control server mode --capacity (default 32) and self-hosted runner capacity cap concurrent sessions. Sources: claude-code:cc-workflow-limits-chk1, claude-code:remote-control, claude-code:selfhost.
Queues [partial] Self-hosted environments queue sessions for runners. Sources: claude-code:selfhost.
Cpu ram [unknown] RAM-aware queues are not established. CPU-bound workflow concurrency is recorded under concurrency and host_awareness.
Backpressure [unknown] No host-level backpressure is documented. Usage-limit waits are documented separately: interactive sessions wait and continue after a claude.ai limit resets. Sources: claude-code:interactive.
Host awareness [partial] Workflow concurrency is CPU-aware: fewer agents run when fewer CPUs are available, including inside a CPU-limited container (workflows). Sources: claude-code:cc-workflow-limits-chk1.
Check reuse [unknown] Check reuse: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Check deduplication [unknown] Check deduplication: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Bounded tests [partial] Foreground Bash waits two minutes by default, up to ten, and moves timed-out commands to the background except sleep-prefixed commands. Unattended background commands have their own time limit; this is not a bound on total test work. Sources: claude-code:cloud-env-nov07.
Cancellation [supported] Users can interrupt, stop and kill background tasks. Sources: claude-code:agent-view.
Fairness [unknown] Fairness: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Recovery [unknown] Recovery: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Shared limits [supported] All Claude and Claude Code usage share subscription limits. Sources: claude-code:support-pro-max.
Live multi-human collaboration is limited. Cloud sessions can be shared by link (Team visibility or public for Pro/Max) but recipients see the latest state without real-time updates; multiple terminals on one session interleave a single transcript.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Team/Enterprise: Private or Team visibility; Pro/Max: Private or Public · OS: Not specified · Execution modes: managed-cloud, local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
No presence, co-prompting or comment threads were found for the CLI; no multi-user test was run.
11 atomic answers
Visibility [partial] Cloud sessions can be shared by link; Team plans offer organization-wide visibility; recipient views do not update in real time. Artifacts can be shared with the organization. Sources: claude-code:web, claude-code:artifacts.
Discussion [unknown] Live discussion inside shared CLI or cloud-session conversations was not established.
Comments [unknown] Comment threads inside shared CLI or cloud-session conversations were not established.
Presence [unknown] Multi-person presence inside shared CLI or cloud-session conversations was not established.
Coprompting [unknown] Multi-person co-prompting inside shared CLI or cloud-session conversations was not established.
Ordering [unknown] Ordering of input from multiple people inside shared CLI or cloud-session conversations was not established.
Conflicts [partial] Resuming the same CLI session in two terminals without forking interleaves messages into one transcript. Sources: claude-code:sessions.
Attribution [partial] Claude replies to GitHub review threads under the user account but labeled as Claude Code; shared sessions show the owner display name. Sources: claude-code:web.
Permissions [partial] Shared sessions can require recipients to have repository access. Verification is enabled by default for Team and Enterprise shares and not enabled by default for Pro and Max shares; it is set under Settings > Claude Code > Sharing settings (web page). Sources: claude-code:web.
Plan restrictions [supported] Team/Enterprise sharing is Private or Team; Pro/Max is Private or Public. Sources: claude-code:web.
Owner online [unknown] Whether the owner must remain online for shared CLI or cloud-session conversation collaboration was not established.
Roles and access are managed at the Claude account level: Owners and admins configure cloud environments and settings; Team adds SSO and central billing; Enterprise adds role-based access, SCIM, audit logs and spend limits.
Scope
Surfaces: Claude Code CLI, Claude Code desktop, Claude Code web/mobile, Claude Code IDE extensions · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Team, Enterprise · OS: Not specified · Execution modes: managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Project-level and task-level roles were not found; role definitions were not read in detail.
9 atomic answers
Organization roles [partial] Owners configure shared cloud environments and defaults; admins govern settings; Enterprise lists role-based access with fine-grained permissioning. Sources: claude-code:admin, claude-code:pricing.
Project roles [unknown] Project roles: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Task roles [unknown] Task roles: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Sso [supported] SSO is listed for Team and Enterprise; Claude apps gateway sign-in uses corporate SSO. Sources: claude-code:pricing, claude-code:auth.
Provisioning [partial] SCIM is listed for Enterprise and configured at the Claude account level; seat assignment is separate. Sources: claude-code:pricing, claude-code:admin.
Unauthorized access [partial] When repository-access verification is enabled in Sharing settings, it limits shared-session recipients. Sources: claude-code:web.
Revocation [unknown] Revocation: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Plan scope [supported] Pricing page separates Team (SSO, central billing) and Enterprise (RBAC, SCIM, audit logs, spend limits, IP allowlisting). Sources: claude-code:pricing.
Approvals are per-user permission prompts governed by permission modes plus allow/ask/deny rules, hooks and organization managed settings. There is no separate approver identity; mobile and Remote Control sessions can answer prompts.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Permission behaviour was read from documentation and not exercised.
Network [partial] Sandboxed commands reach only allowed domains and WebFetch has domain permissions; other hosts follow the permission mode. Sources: claude-code:sandboxing, claude-code:permissions.
Secret [unknown] Secret: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Delivery [partial] Prompts appear in the terminal; PermissionRequest and Notification hooks and mobile push can relay them. Sources: claude-code:hooks, claude-code:mobile.
Approver identity [partial] The approver is the local session user; shared approver roles were not described. Sources: claude-code:permissions.
Lifetime [unknown] Lifetime: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Deny [supported] Deny rules block in every mode, including bypassPermissions. Sources: claude-code:permission-modes.
Revoke [unknown] Revoke: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Tunnels [unknown] Tunnels: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Defaults [supported] Auto mode (classifier) is the built-in starting mode for interactive terminal and VS Code sessions from v2.1.283. Sources: claude-code:permission-modes.
Enforcement [supported] Managed settings take precedence over local settings, and bypass can be restricted. Sources: claude-code:permissions, claude-code:admin.
History comes from local session transcripts, checkpoints, OpenTelemetry events and, on Enterprise, audit logs. CLI transcripts default to 30 days of local retention, and Desktop and Cowork sessions are exempt by default (data-usage page); cross-user persistence and integrity guarantees were not found.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Enterprise audit logs · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Approvals [unknown] Approvals: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Discussion [unknown] Discussion: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Decisions [unknown] Decisions: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Cross user persistence [unknown] Cross user persistence: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Cross session persistence [supported] Sessions are saved and resumed across process restarts. Sources: claude-code:sessions.
Mutability [partial] Local transcripts are plaintext files under ~/.claude/projects. Sources: claude-code:data-usage.
Retention [supported] Local CLI transcripts are kept 30 days by default (cleanupPeriodDays). Transcripts of sessions started or most recently continued in Claude Desktop or Cowork are exempt from the 30-day cleanup by default (data-usage page). Commercial server retention is 30 days standard. Sources: claude-code:data-usage, claude-code:data-usage.
Integrity [unknown] Integrity: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Review is available through /diff, Desktop diff view with inline comments, local /code-review, and managed GitHub PR Code Review (Team/Enterprise) whose findings do not block merges. The Team and Enterprise review feature is not available for organizations with Zero Data Retention or the HIPAA configuration (code-review page). Batch review UX was not established.
Scope
Surfaces: Claude Code CLI, Claude Code desktop · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
Not separately cross-checked
No review experience test; effort, keyboard and narrow-screen behaviour were not examined.
Managed Code Review also excludes organizations using customer-managed encryption keys; the current page no longer labels it research preview.
Change understanding [unknown] Change understanding: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Evidence navigation [unknown] Evidence navigation: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Request changes [partial] Code Review findings are tagged by severity and do not approve or block PRs. Sources: claude-code:code-review-nov07.
Shipping guidance [partial] Review focus can be tuned with CLAUDE.md or REVIEW.md. Sources: claude-code:code-review-nov07.
Effort [unknown] Effort: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Missed evidence [unknown] Missed evidence: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Keyboard [unknown] Keyboard: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Narrow screen [unknown] Narrow screen: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Context switching [unknown] Context switching: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Roles [unknown] Roles: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Retained context [unknown] Retained context: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Takeover is mostly same-user: teleport and Remote Control move a session between one person own devices; colleagues can open shared cloud session links to see the latest state. Transfer of ownership between users was not found.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: managed-cloud, remote-control · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
No colleague takeover scenario was run.
12 atomic answers
Ongoing takeover [unknown] Ongoing takeover: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Same task [unknown] Same task: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Session history [supported] Sessions can be resumed from local transcripts or reopened from claude.ai/code with history restored. Sources: claude-code:sessions, claude-code:web.
Decisions [unknown] Decisions: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Artifacts [unknown] Artifacts: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Permissions [partial] Teleport requires the same claude.ai account and a checkout of the same repository; shared links can require repository access when that setting is enabled. Sources: claude-code:web.
Runtime context [partial] Project CLAUDE.md is shared via the repository; auto memory stays on the individual machine. Sources: claude-code:memory.
Owner disconnect [supported] Cloud sessions continue after the owner disconnects; local sessions and Remote Control need the machine on. Sources: claude-code:web, claude-code:mobile.
Ownership [unknown] Ownership: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Attribution [unknown] Attribution: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Context loss [unknown] Context loss: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Harness transfer [unsupported] All workers are Claude sessions; no cross-harness transfer is offered. Sources: claude-code:agents.
Outputs include diffs (/diff, Desktop diff view), live Artifact pages published to claude.ai, Desktop app preview, and session links to PRs. An artifact gallery lists created artifacts. Retention of test outputs and recordings was not established.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Artifacts: Pro, Max, Team, Enterprise · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Artifact availability requires a claude.ai login and plan; no output retention test was run.
Documents [supported] Artifacts publish interactive pages that can be shared privately, with the organization or publicly. Sources: claude-code:artifacts.
Screenshots [unknown] Screenshots: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Recordings [unknown] Recordings: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Previews [partial] Desktop previews the running app in a Browser pane. Sources: claude-code:desktop.
Test outputs [unknown] Test outputs: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Reviewer outputs [unknown] Reviewer outputs: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Gallery [supported] A gallery of artifacts at claude.ai/code/artifacts lists them; the artifact page header links to it. Sources: claude-code:artifacts.
Revision links [unknown] Revision links: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Persistence [partial] Artifacts update in place as the session continues and are versioned for sharing. Sources: claude-code:artifacts.
Child findings [unknown] Child findings: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Prompt leakage [unknown] Prompt leakage: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Verification is policy you configure: hooks run commands and can block actions or task completion (exit code 2), /verify builds and drives the app, /goal re-checks a completion condition, and /code-review and managed Code Review provide independent review. Managed settings can restrict which hooks run (allowManagedHooksOnly), but the docs do not state that a check must pass before delivery.
Scope
Surfaces: Claude Code CLI · Version: Current first-party documentation read 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Lint [unknown] Lint: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Types [unknown] Types: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Independent review [supported] /code-review reviews the current diff or a PR for correctness bugs (ultra runs a deep cloud review); managed Code Review (Team/Enterprise) reviews GitHub PRs with several agents in parallel and a verification step. Sources: claude-code:commands, claude-code:code-review-nov07.
Spec checks [unknown] Spec checks: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Required checks [unknown] Required checks: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Failure blocking [supported] Blocking depends on the hook event. TaskCompleted exit code 2 prevents task completion and TeammateIdle exit code 2 keeps the teammate working; SessionStart exit code 2 does not block execution. Sources: claude-code:hooks, claude-code:teams.
Requirement conflicts [unknown] Requirement conflicts: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Configured policy [supported] Hooks are configured in settings or plugins. Sources: claude-code:hooks.
Enforced policy [partial] Managed settings take precedence; customization lockdown can restrict hooks to managed sources. Sources: claude-code:admin, claude-code:permissions.
The CLI supports git workflows: it can create commits and pull requests on request, run in isolated worktrees, link sessions to PRs, and (cloud) auto-fix CI failures and review comments on GitHub.
Scope
Surfaces: Claude Code CLI, Claude Code web/mobile · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
generally-available
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
GitHub is the supported PR host for cloud features.
Comments [partial] Cloud auto-fix replies to review threads on GitHub labeled as Claude Code; Code Review posts inline comments. Sources: claude-code:web, claude-code:code-review.
Ci response [partial] Cloud auto-fix investigates CI check failures and pushes fixes; it needs the Claude GitHub App. Sources: claude-code:web.
Rebase [partial] Auto-fix cannot detect merge conflicts; users ask Claude to rebase. Sources: claude-code:web.
Conflicts [partial] Conflict resolution is manual-prompted. Sources: claude-code:web.
Product state [unknown] Product state: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Merge and deployment authority is delegated to git/CLI permissions: Bash allow, ask or deny rules can gate push and merge commands. Dedicated approval, release and deploy workflows were not found.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
No merge or deploy flow was tested.
6 atomic answers
Approve [unknown] Approve: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Merge [unknown] Merge: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Release [unknown] Release: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Deploy [unknown] Deploy: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Exact revision [unknown] Exact revision: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Authority enforcement [partial] Permission rules on Bash command patterns (for example git push patterns) can allow, ask or deny actions. Sources: claude-code:permissions.
Completion evidence is session links to PRs and OTel commit/PR counts; rollback is via /rewind checkpoints on local files. Build and deployment links and stale-approval invalidation were not found.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Local rewind is not a release rollback.
9 atomic answers
Commit links [partial] OTel commit and pull request counters record counts of commits and PRs created, not links to them. Sources: claude-code:monitoring.
Pr links [supported] --from-pr opens the session picker filtered to sessions linked to a pull request. Sources: claude-code:sessions.
Build links [unknown] Build links: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Deployment links [unknown] Deployment links: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Delivery status [unknown] Delivery status: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Retained evidence [unknown] Retained evidence: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Rollback [partial] /rewind restores code and conversation to a checkpoint; this is not a deployment rollback. Sources: claude-code:checkpointing.
Revert [unknown] Revert: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Stale approval invalidation [unknown] Stale approval invalidation: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
High-volume shipping features are limited: agent view summaries, /batch splitting a change into worktree subagents, and Code Review severity tags. A review queue with exact-revision approval was not found.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: local · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Not a review queue.
12 atomic answers
Review queue [unknown] Review queue: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Summaries [partial] Agent view rows show each session latest status and one-line report. Sources: claude-code:agent-view.
Grouping [unknown] /batch splits one large change into 5 to 30 worktree subagents; that is task splitting, not a review grouping. Sources: claude-code:agents.
Risk size filters [unknown] Risk size filters: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Findings [partial] Code Review findings are tagged by severity. Sources: claude-code:code-review.
Artifacts [unknown] Artifacts: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Request changes [unknown] Request changes: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Exact revision approval [unknown] Exact revision approval: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Dependency queues [unknown] Dependency queues: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Next actions [unknown] Next actions: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Stale approval invalidation [unknown] Stale approval invalidation: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Check deduplication [unknown] Check deduplication: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: Not specified · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
No source-inspected evidence recorded.
Head commit 71cdddec623889d38af14b7a489670a03186f659 was not re-verified on 2026-10-09; the repository main page did not expose it.
Server license [unknown] Server license: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Hosted extensions [partial] The Python Agent SDK repository (MIT per GitHub) and sandbox-runtime (Apache-2.0) are open source; hosted-service licensing was not separately established. Sources: claude-code:sdk-repo, claude-code:sandbox-repo.
Source visibility [partial] A public GitHub repository exists for issues, changelog, plugins and examples; the CLI source was not inspected. Sources: claude-code:repo.
Source revision [partial] Latest release is v2.1.295 (2026-10-08); its tag resolves to commit 602df92 (release page). v2.1.294 is the prior release. The stored head commit 71cdddec623889d38af14b7a489670a03186f659 is not shown on the release page and was not re-verified on 2026-10-09; no source file was inspected. Sources: claude-code:release-2-1-295, claude-code:repo, claude-code:changelog.
Hosted parity [unknown] Hosted parity: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
The installed Claude Code CLI can run its full agent loop, tools and context management on user-controlled infrastructure, including scripts and CI/CD. Model inference/authentication still uses Anthropic or a supported provider. This does not self-host the Claude web/desktop cloud control plane; cloud self-hosted runners are a separate worker feature.
Scope
Surfaces: Claude Code CLI · Version: Claude Code CLI headless documentation checked 2026-11-13 for local-core self-hosting. Other atomic answers retain their individually dated source scope. · Plan: Not specified · OS: Not specified · Execution modes: user-controlled CLI core · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-13 (9 days old)
Agent cross-check
Not separately cross-checked
The mapping covers the local CLI core only; hosted cloud/desktop service self-hosting is not established.
8 atomic answers
Self hosting [supported] The CLI runs the full Claude Code agent loop and tools locally through claude -p, including user-managed scripts and CI/CD. Sources: claude-code:gap-local-core.
Vendor dependencies [supported] Requires Anthropic API or a supported cloud provider and internet access; cloud features need claude.ai. Sources: claude-code:setup, claude-code:feature-availability.
Import [partial] /import brings configuration from Codex, Gemini CLI or Cursor; session-history import was not described. Sources: claude-code:commands.
Cancellation [supported] Cancellation (pricing FAQ, retrieved 2026-10-09): cancel at least 24 hours before renewal; the plan stays active to the end of the billing period. Web or desktop subscriptions cancel under Settings > Billing; App Store or Google Play subscriptions cancel in that store. Team and self-serve Enterprise Owners cancel under Organization settings > Billing; sales-assisted Enterprise cancels through its Anthropic contact or sales. Sources: claude-code:pricing.
Account independence [partial] Console API keys and Bedrock/Agent Platform/Foundry credentials work without a claude.ai subscription for the CLI. Sources: claude-code:feature-availability.
History portability [unknown] History portability: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Prompts and outputs go to Anthropic (or the chosen provider) over TLS 1.2+; consumer plans train on data only when the setting is on, commercial plans do not by default; standard retention is 30 days for commercial and 30 days or 5 years for consumers. Usage metrics default on for the Claude API and off on Amazon Bedrock, Google Cloud's Agent Platform and Microsoft Foundry; error reports default on only for Pro, Max and Ultra sign-ins on v2.1.198+; both have documented opt-outs.
Scope
Surfaces: Claude Code CLI · Version: Current first-party documentation read 2026-11-07; changelog through v2.1.319 · Plan: Retention and training differ by plan · OS: Not specified · Execution modes: local, managed-cloud · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Plan-specific; Remote Control and cloud flows differ.
15 atomic answers
Code data [supported] Local files stay on the machine; code in prompts and tool results is sent to the model API; cloud sessions clone repositories into a VM. Sources: claude-code:data-usage-second-review-nov07.
Artifacts [unknown] Artifacts: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Retention [supported] Consumer: 5 years if training allowed, else 30 days; commercial: 30 days, ZDR for qualified Enterprise. Sources: claude-code:data-usage-second-review-nov07.
Subprocessors [unknown] Subprocessors: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Telemetry defaults [supported] Metrics default on for the Claude API; error reports on for Pro/Max/Ultra sign-ins on v2.1.198+; both off by default on Bedrock, Foundry, Agent Platform. Sources: claude-code:data-usage-second-review-nov07.
Consent prompt [unknown] First-run telemetry prompts were not observed. The documented session-survey follow-up asks before any transcript upload; that is a separate consent step, not a telemetry prompt. Sources: claude-code:data-usage-second-review-nov07.
Provider separation [supported] Third-party providers default telemetry off, except that hosts setting CLAUDE_CODE_PROVIDER_MANAGED_BY_HOST default metrics on. Sources: claude-code:data-usage-second-review-nov07.
Pro $20/month or $200/year; Max $100/month; Ultra $200/month. Team seats $25/$125 monthly or $20/$100 monthly-equivalent annual. Enterprise $20/seat/month annual with 20-seat minimum, plus API usage. API usage is billed at list prices per million tokens: Fable 5.2 $8/$40, Opus 5.5 $4/$20, Sonnet 5.5 $2/$10, Haiku 5.5 $0.10/$0.50 up to 100K-token prompts ($0.50/$2.50 above). Batch saves 50%; US-only inference costs 1.1x; fast mode costs 2x. Trials appear only in the Team/Enterprise plan comparison: No for Team and self-serve Enterprise, Yes for sales-assisted Enterprise; trial length and terms are not stated. The page states no trial for the Free, Pro, Max or Ultra plans, whose comparison has no Trials row.
Scope
Surfaces: Claude Code CLI, Claude Code desktop, Claude Code web/mobile, Claude Code IDE extensions · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Pro, Max, Ultra, Team, Enterprise · OS: Not specified · Execution modes: Not specified · Region: Not specified · Harness: Not specified
Availability
generally-available
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-21
No workload, billing or authenticated client test was performed; unestablished details remain unknown.
Credits [supported] Max/Ultra monthly platform API credits cannot be used for Claude Code. Sources: claude-code:support-pro-max-nov07.
Trial [partial] Trials appear only in the Team/Enterprise plan comparison: No for Team and self-serve Enterprise, Yes for sales-assisted Enterprise; trial length and terms are not stated. The page states no trial for the Free, Pro, Max or Ultra plans, whose comparison has no Trials row. Sources: claude-code:pricing-nov21.
Model cost [supported] API usage is billed at list prices per million tokens: Fable 5.2 $8/$40, Opus 5.5 $4/$20, Sonnet 5.5 $2/$10, Haiku 5.5 $0.10/$0.50 up to 100K-token prompts ($0.50/$2.50 above). Batch saves 50%; US-only inference costs 1.1x; fast mode costs 2x. Sources: claude-code:pricing-nov21.
Compute cost [partial] Cloud sessions have no separate compute charge under the documented subscription route. Sources: claude-code:web-nov07.
Fixed subscription examples: one Pro $20/month or $200/year; two Team Standard seats $50/month or $480/year; five $125/month or $1,200/year. Variable workload costs remain unknown.
Scope
Surfaces: Claude Code CLI · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: Not specified · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
No workload, billing or authenticated client test was performed; unestablished details remain unknown.
11 atomic answers
One person [partial] One Pro subscription $20/month or $200/year. This is a seat baseline, not a measured workload. Sources: claude-code:pricing-nov07.
Two people [partial] Two Standard Team seats $50/month, or $480/year. Sources: claude-code:pricing-nov07.
Five people [partial] Five Standard Team seats $125/month, or $1,200/year. Sources: claude-code:pricing-nov07.
Task assumptions [unknown] Not established by the first-party sources reread November 7; no configured or authenticated scenario was tested.
Concurrency [unknown] Not established by the first-party sources reread November 7; no configured or authenticated scenario was tested.
Tokens [unknown] Not established by the first-party sources reread November 7; no configured or authenticated scenario was tested.
Compute hours [unknown] Not established by the first-party sources reread November 7; no configured or authenticated scenario was tested.
Fixed cost [partial] Standard Team $25 monthly/$20 annual-equivalent; Premium and Enterprise have different rates/minimums. Sources: claude-code:pricing-nov07.
GitHub releases run to v2.1.322 (Nov 10, 2026, marked Latest). Feature preview/beta flags, country/account access and status history are documented separately; no response-time guarantee was established.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.322 (releases retrieved 2026-11-12) · Plan: Not specified · OS: Not specified · Execution modes: Not specified · Region: Not specified · Harness: Not specified
Response commitments [unknown] Response commitments: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Release activity [supported] v2.1.319 (Nov 6), v2.1.321 (Nov 9) and v2.1.322 (Nov 10, 2026, marked Latest); no binary was run. Sources: claude-code:nov12-releases.
Status history [partial] Status page and API provide component/incident reports; Code and Console are operational at retrieval, not a reliability guarantee. Sources: claude-code:status-nov07, claude-code:status-summary-nov07.
Deprecation [partial] The README marks npm installation as deprecated in favor of native install Sources: claude-code:repo-nov07.
Export options [unknown] Export options: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Claude Code documentation points to the Anthropic Trust Center for compliance artifacts. The Trust Center was not fetched; SOC 2 Type 2 and ISO 27001 scope, issuer, validity and whether they cover Claude Code were not verified.
Scope
Surfaces: Claude Code CLI · Version: Claude Code v2.1.295 (released 2026-10-08); official docs and pricing page read 2026-10-09 · Plan: Not specified · OS: Not specified · Execution modes: Not specified · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-10-09 (44 days old)
Agent cross-check
2026-11-07
Vendor-stated; no certificate or report read.
9 atomic answers
Iso scope [partial] Docs name an ISO 27001 certificate (vendor-stated, not verified). Sources: claude-code:security.
Iso issuer [unknown] Iso issuer: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Iso validity [unknown] Iso validity: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Soc2 type [partial] Docs name a SOC 2 Type 2 report (vendor-stated, not verified). Sources: claude-code:security.
Soc2 scope [unknown] Soc2 scope: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Soc2 period [unknown] Soc2 period: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Report access [partial] The security page links the Trust Center for the SOC 2 Type 2 report and ISO 27001 certificate; whether they cover Claude Code scope was not stated and the Trust Center was not readable. Sources: claude-code:security.
Verification date [unknown] Verification date: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Evidence status [partial] Only first-party mentions were checked; the Trust Center portal is JavaScript-rendered and its contents were not read. Sources: claude-code:security.
Free excludes Claude Code; all paid plans include it. Evaluation requires an eligible subscription, Console/API billing or supported provider route; bundled platform credits do not pay Code usage.
Scope
Surfaces: Claude Code CLI, Claude Code desktop, Claude Code web/mobile, Claude Code IDE extensions · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Free plan excluded · OS: Not specified · Execution modes: Not specified · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Console API credit promotions and trials were not checked.
Console API credit promotions were not checked; trial length and terms for sales-assisted Enterprise are not stated.
10 atomic answers
Ongoing free [unsupported] Pricing page lists Claude Code as No for the Free plan. Sources: claude-code:pricing-nov07.
Trial [partial] Trial terms are not established. The pricing comparison marks Trials No for Team and self-serve Enterprise and Yes only for sales-assisted Enterprise (pricing page). Sources: claude-code:pricing-nov07.
Credits trial [unknown] Credits trial: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Self host cost [unknown] Self host cost: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Paid only [supported] All paid Claude plans include Code; Free does not. Sources: claude-code:pricing-nov07.
Early access [unknown] The pricing page Max line (early access to advanced Claude features) does not name Claude Code; no Claude Code early-access gate was found.
Free scenarios [unknown] Free scenarios: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Card required [unknown] Card required: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Public seat prices, plan multiples and separate API rates are stated. Enterprise has a 20-seat annual minimum; model/task costs vary, and bundled Max/Ultra platform credits exclude Code usage.
Scope
Surfaces: Claude Code CLI, Claude Code desktop, Claude Code web/mobile, Claude Code IDE extensions · Version: First-party public docs reread 2026-11-07; changelog through v2.1.319 · Plan: Not specified · OS: Not specified · Execution modes: Not specified · Region: Not specified · Harness: Not specified
Availability
unknown
Delivery
native
Checked
2026-11-07 (15 days old)
Agent cross-check
2026-11-07
Judgement is the researcher classification, not a vendor statement.
9 atomic answers
Units [partial] Per-seat monthly price plus usage limits; API pricing per million tokens. Sources: claude-code:pricing-nov07.
Compute markup [supported] No separate compute charge for cloud VMs. Sources: claude-code:web-nov07.
Model markup [unknown] Model markup: not established by the first-party docs, pricing page, changelog and repository checked on 2026-10-09; no hands-on test was run.
Minimum commitment [supported] Enterprise is annual with 20-seat minimum; Team starts at two seats. Sources: claude-code:pricing-nov07.
Cancellation [supported] Cancellation (pricing FAQ, read 2026-11-07): cancel at least 24 hours before renewal; the plan stays active to the end of the billing period. Web or desktop subscriptions cancel under Settings > Billing; App Store or Google Play subscriptions cancel in that store. Team and self-serve Enterprise Owners cancel under Organization settings > Billing; sales-assisted Enterprise cancels through its Anthropic contact or sales. Sources: claude-code:pricing-nov07.
Documents terminal (macOS/Linux/WSL/Windows), VS Code, desktop (macOS, Windows x64/ARM64, Ubuntu/Debian beta), web and JetBrains surfaces; most surfaces need a Claude subscription or Console account; CLI, VS Code and JetBrains also support third-party providers.
Kind
documented
Locator
Get started tabs; Terminal install; Desktop app; Web
Compares CLI, Desktop, VS Code, JetBrains, Web and Mobile; CLI is the most complete surface (scripting and Agent SDK are CLI-only); Dispatch, Remote Control, Channels, Slack, self-hosted environments and scheduled tasks listed with where Claude runs and plan requirements.
Kind
documented
Locator
Where to run Claude Code; Work when you are away from your terminal
The Claude iOS/Android app is a client for cloud sessions, projects, Remote Control and Dispatch (no code runs on the phone). Push notifications when Remote Control is active; local-only commands such as /plugin and /resume do not work; Bypass permissions cannot be selected from the app.
Kind
documented
Locator
Work from your phone; Get push notifications; Limitations
Drive a local CLI/Desktop/VS Code session from claude.ai/code or the mobile app; machine must stay on; Pro/Max/Team/Enterprise only (no API keys, Bedrock/Vertex/Foundry, gateways); server mode default capacity 32; transcript stored on Anthropic servers while connected.
Kind
documented
Locator
Requirements; Start a Remote Control session; Limitations; Connection and security
Desktop Code tab has parallel sessions with Git isolation, diff review and comments, app preview, integrated terminal, permission modes, scheduled tasks and Dispatch from phone; needs a paid subscription; Desktop does not accept a Console API key directly.
Kind
documented
Locator
Sessions; Choose a permission mode; Work in parallel; Dispatch; Coming from the CLI
Linux desktop app is beta on Ubuntu 22.04+/Debian 12+; Fedora/Arch should use the CLI; signs in with claude.ai subscription or SSO, not a Console API key.
VS Code 1.94.0+; extension version equals bundled Claude Code version; inline diffs, plan review, permission mode indicator; also installable in Cursor; shares ~/.claude settings with the CLI.
Plugin runs the separately installed claude CLI in the IDE terminal and adds diff viewing, selection and diagnostic sharing; it does not bundle the CLI.
Cloud sessions on Pro, Max, Team and Enterprise (premium seats); Anthropic-managed VM by default; clone/PR creation requires GitHub (GHES for Team/Enterprise); non-GitHub repos via local bundle <100 MB; idle sessions pause and VMs are reclaimed, conversation history restored on reopen; shared rate limits, no separate compute charge; CLI --cloud and --teleport (one-way cloud to terminal); auto-fix of PR CI failures and review comments needs the Claude GitHub App.
OS-enforced sandbox for shell commands on macOS, Linux and WSL2 (off by default; native Windows runs unsandboxed); built on open source sandbox-runtime; reads allowed broadly by default, writes limited to working directory, network via domain-allowlist proxy; file tools, MCP servers and hooks run outside it.
Kind
documented
Locator
What the sandbox restricts; What runs outside the sandbox
Hooks fire on events such as PreToolUse, PostToolUse, PermissionRequest, Notification, Stop, SubagentStop, SessionStart/End, TaskCreated/Completed, WorktreeCreate and can be commands, prompts, agents or HTTP calls; can block or auto-approve actions.
Kind
documented
Locator
What you can automate; Hook types; Hooks and permission modes
Local CLI features work across supported providers. Desktop is an exception: gateway/3P deployments can run Code on Bedrock, Google Agent Platform, Foundry, self-hosted gateways and the Claude API with Console key. Subscription-only features have separate plan restrictions.
Kind
documented
Locator
Availability by model provider; Features that require a Claude subscription
Remote Control connects claude.ai/code or the mobile apps to a local Claude session. Users can send messages from the terminal, browser and phone; the local host must remain running. Eligible Claude subscriptions are required.
Kind
documented
Locator
Introduction; Connect from another device; Requirements
Account types: Claude Pro/Max, Teams/Enterprise, Claude Console, Bedrock/Vertex(Agent Platform)/Foundry, Claude apps gateway. Multiple accounts via separate CLAUDE_CONFIG_DIR directories. ANTHROPIC_API_KEY approval skips login.
Kind
documented
Locator
Log in to Claude Code; Log in with multiple accounts
Python and TypeScript libraries run the Claude Code agent loop with built-in tools, hooks, subagents, MCP, permissions and sessions; non-Python/TS languages can drive the CLI as a subprocess.
Kind
documented
Locator
Compare the Agent SDK to other Claude tools; Capabilities
Subagents, agent view (research preview), agent teams (experimental), dynamic workflows, and Projects. In every approach the workers are Claude sessions; other tools are exposed via MCP. Worktrees isolate parallel sessions.
Kind
documented
Locator
Five ways to work on several tasks; Choose an approach; Check on running work
MCP client supports http, sse, stdio and websocket servers with OAuth 2.0, local/project/user scopes and claude.ai connectors; claude mcp serve exposes Claude Code as a stdio MCP server; channels push events.
Kind
documented
Locator
Installing MCP servers; Authenticate with remote MCP servers; Use Claude Code as an MCP server; Scopes
Research preview. MCP-based channels push Telegram/Discord/iMessage or webhook events into an open session; requires claude.ai login or Console key; not on Bedrock/Agent Platform/Foundry; Team/Enterprise must enable.
/usage shows client-side session cost estimate at list price, plan usage bars and attribution to skills/subagents/plugins/MCP; usage credits past plan limits; --max-budget-usd and --max-turns in print mode; org spend limits for Team/Enterprise; average enterprise cost about $13 per developer per active day.
Kind
documented
Locator
Track your costs; Plan usage breakdown; Usage credits; Manage costs for your organization
Current list prices: Pro $20 monthly or $200/year displayed as $17/month; Max $100 monthly; Ultra $200 monthly. Team 2-150 users: Standard $25 monthly/$20 annual per month, Premium $125/$100. Enterprise $20/seat/month annual, 20-seat minimum, plus API usage. All paid plans include Code; taxes excluded. Max/Ultra platform API credits cannot pay Code usage.
Kind
documented
Locator
Individual and Team and Enterprise plan cards; feature comparison
Pro, Max and Ultra include Code in a shared subscription quota; Max and Ultra platform API credits cannot pay Code usage. ANTHROPIC_API_KEY causes API billing instead of subscription. No workload usage test.
Documents terminal (macOS/Linux/WSL/Windows), VS Code, desktop (macOS, Windows x64/ARM64, Ubuntu/Debian beta), web and JetBrains surfaces; most surfaces need a Claude subscription or Console account; CLI, VS Code and JetBrains also support third-party providers.
Kind
documented
Locator
Get started tabs; Terminal install; Desktop app; Web
Account types: Claude Pro/Max, Teams/Enterprise, Claude Console, Bedrock/Vertex(Agent Platform)/Foundry, Claude apps gateway. Multiple accounts via separate CLAUDE_CONFIG_DIR directories. ANTHROPIC_API_KEY approval skips login.
Kind
documented
Locator
Log in to Claude Code; Log in with multiple accounts
Drive a local CLI/Desktop/VS Code session from claude.ai/code or the mobile app; machine must stay on; Pro/Max/Team/Enterprise only (no API keys, Bedrock/Vertex/Foundry, gateways); server mode default capacity 32; transcript stored on Anthropic servers while connected.
Kind
documented
Locator
Requirements; Start a Remote Control session; Limitations; Connection and security
Cloud sessions on Pro, Max, Team and Enterprise (premium seats); Anthropic-managed VM by default; clone/PR creation requires GitHub (GHES for Team/Enterprise); non-GitHub repos via local bundle <100 MB; idle sessions pause and VMs are reclaimed, conversation history restored on reopen; shared rate limits, no separate compute charge; CLI --cloud and --teleport (one-way cloud to terminal); auto-fix of PR CI failures and review comments needs the Claude GitHub App.
Public beta on Team and Enterprise (off by default); runners on customer infrastructure execute cloud sessions; unavailable with ZDR or HIPAA; model inference still via Anthropic API unless Bedrock/Agent Platform configured; session content goes to Anthropic.
Kind
documented
Locator
How self-hosted environments work; Availability and limitations
Research preview. claude agents dispatches and monitors background local sessions; sessions move into git worktrees before editing; background sessions use subscription quota like interactive ones, stop if the machine shuts down; claude agents --json output.
Kind
documented
Locator
Quick start; How file edits are isolated; Limitations; Version history
claude --worktree creates .claude/worktrees/<name> on a new branch; requires a git repository (hooks can replace git logic); subagents can use isolation: worktree.
Kind
documented
Locator
Start Claude in a worktree; Isolate subagents; Non-git version control
Weekly digests list cross-session messaging and self-hosted environments (Week 32, public beta), auto mode default (from August 14 for Pro/Max/Team), Desktop resume of CLI sessions (Week 35), Fable 5.1 (Week 36); no future-dated commitments found.
Anthropic-hosted sessions use fresh Ubuntu 24.04 x86_64 VMs with approximate ceilings of 4 vCPUs, 16 GB RAM and 30 GB disk. Setup under roughly five minutes is cached as a filesystem snapshot, expiring after roughly seven days. Foreground Bash defaults to two minutes, up to ten; timed-out commands move to background except sleep-prefixed commands. Unattended background commands have a separate time limit. Idle VMs pause and can be reclaimed. Network secrets are available on Pro, Max, Ultra, Team and Enterprise, with role/configuration requirements.
Kind
documented
Locator
Installed tools; Time limits; Setup scripts; Environment caching; Network access
Cloud routines are now named cloud scheduled tasks; shared web/Desktop/CLI account records, schedule/API/GitHub triggers, research preview fire endpoint, and per-account execution. /schedule creates them and /routines remains an alias.
Kind
documented
Locator
Triggers; Add an API trigger; GitHub event; Usage and limits
Research preview. claude agents dispatches and monitors background local sessions; sessions move into git worktrees before editing; background sessions use subscription quota like interactive ones, stop if the machine shuts down; claude agents --json output.
Kind
documented
Locator
Quick start; How file edits are isolated; Limitations; Version history
Documents claude agents, claude remote-control, claude mcp, claude gateway; flags --worktree, --max-turns and --max-budget-usd (print mode only), --fallback-model, --output-format, --permission-mode, --bare, --channels.
Python and TypeScript libraries run the Claude Code agent loop with built-in tools, hooks, subagents, MCP, permissions and sessions; non-Python/TS languages can drive the CLI as a subprocess.
Kind
documented
Locator
Compare the Agent SDK to other Claude tools; Capabilities
Hooks fire on events such as PreToolUse, PostToolUse, PermissionRequest, Notification, Stop, SubagentStop, SessionStart/End, TaskCreated/Completed, WorktreeCreate and can be commands, prompts, agents or HTTP calls; can block or auto-approve actions.
Kind
documented
Locator
What you can automate; Hook types; Hooks and permission modes
CLAUDE.md at managed/user/project/local scope with imports and .claude/rules; AGENTS.md read when no CLAUDE.md (v2.1.277+); auto memory stored per project under ~/.claude/projects/<project>/memory/ as editable markdown; /memory browses and toggles.
Kind
documented
Locator
CLAUDE.md files; AGENTS.md; Auto memory; View and edit with /memory
MCP client supports http, sse, stdio and websocket servers with OAuth 2.0, local/project/user scopes and claude.ai connectors; claude mcp serve exposes Claude Code as a stdio MCP server; channels push events.
Kind
documented
Locator
Installing MCP servers; Authenticate with remote MCP servers; Use Claude Code as an MCP server; Scopes
Managed settings delivered via claude.ai admin console, MDM or file; SSO, SCIM and seat assignment at the Claude account level; Owners govern cloud environments and the default environment; org effort limits enforced server-side.
Kind
documented
Locator
API providers; managed settings; usage monitoring; cloud sessions admin
Documents claude agents, claude remote-control, claude mcp, claude gateway; flags --worktree, --max-turns and --max-budget-usd (print mode only), --fallback-model, --output-format, --permission-mode, --bare, --channels.
Compares CLI, Desktop, VS Code, JetBrains, Web and Mobile; CLI is the most complete surface (scripting and Agent SDK are CLI-only); Dispatch, Remote Control, Channels, Slack, self-hosted environments and scheduled tasks listed with where Claude runs and plan requirements.
Kind
documented
Locator
Where to run Claude Code; Work when you are away from your terminal
The Claude iOS/Android app is a client for cloud sessions, projects, Remote Control and Dispatch (no code runs on the phone). Push notifications when Remote Control is active; local-only commands such as /plugin and /resume do not work; Bypass permissions cannot be selected from the app.
Kind
documented
Locator
Work from your phone; Get push notifications; Limitations
Desktop Code tab has parallel sessions with Git isolation, diff review and comments, app preview, integrated terminal, permission modes, scheduled tasks and Dispatch from phone; needs a paid subscription; Desktop does not accept a Console API key directly.
Kind
documented
Locator
Sessions; Choose a permission mode; Work in parallel; Dispatch; Coming from the CLI
VS Code 1.94.0+; extension version equals bundled Claude Code version; inline diffs, plan review, permission mode indicator; also installable in Cursor; shares ~/.claude settings with the CLI.
Plugin runs the separately installed claude CLI in the IDE terminal and adds diff viewing, selection and diagnostic sharing; it does not bundle the CLI.
Research preview. Cloud routines with schedule, API (/fire endpoint under experimental beta header, claude.ai users only) and GitHub event triggers; Pro/Max/Team/Enterprise; run on Anthropic cloud or self-hosted environment; belong to individual account; hourly/webhook caps apply.
Kind
documented
Locator
Triggers; Add an API trigger; GitHub event; Usage and limits
The earlier Slack bot is available on Pro, Max and Ultra only in workspaces not connected to an organization Claude Tag. Team/Enterprise use Claude Tag. The retirement notice effective October 5 applies in an organization-connected workspace using the earlier bot.
Product Task tools create, read, update and delete tasks. They are model-dependent or opt-in since v2.1.268, and provided across models in cloud/background sessions.
Task checklists are distinct from running shells/subagents, retain state through compaction and resume, and can share a named on-disk task directory across sessions.
Experimental, disabled by default (CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1). Lead plus independent teammates with own context, direct messaging, shared task list with pending/in-progress/completed states and task dependencies, file-locked claiming; limitations include no in-process teammate resume, one team per session, status lag.
Kind
documented
Locator
Enablement; shared task list; Assign and claim tasks; Limitations
/loop and cron tools are session-scoped (1 minute minimum, recurring tasks expire after 7 days, only fire while Claude Code is running and idle, no catch-up for missed fires); cloud routines minimum 1 hour; Desktop tasks need machine on.
Modes default/Manual, acceptEdits, plan, auto (classifier), dontAsk, bypassPermissions; deny rules apply in every mode; auto is the built-in starting mode for interactive terminal and VS Code sessions from v2.1.283.
Sessions saved continuously as local JSONL transcripts; --continue, --resume, --from-pr, /branch; running the same session in two terminals interleaves one transcript; /export writes plain text.
plansDirectory sets where plan mode stores plan files, resolved relative to the project root; unset defaults to ~/.claude/plans, and paths outside the project root fall back to ~/.claude/plans.
Experimental, disabled by default (CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1). Lead plus independent teammates with own context, direct messaging, shared task list with pending/in-progress/completed states and task dependencies, file-locked claiming; limitations include no in-process teammate resume, one team per session, status lag.
Kind
documented
Locator
Enablement; shared task list; Assign and claim tasks; Limitations
Lists built-in tools including TaskCreate/TaskUpdate/TaskList, TaskStop, WebFetch, WebSearch, Workflow; task tools are provided by default only on some model families.
Kind
documented
Locator
Tool list; Task tool availability; WebFetch and WebSearch
Finding symbols can cost many file reads and grep calls; LSP code intelligence plugins add definition/reference lookup; an organization code search or RAG index can be exposed as an MCP tool.
Kind
documented
Locator
Code intelligence plugins; Expose your own index over MCP
Claude can create commits and pull requests on request; --from-pr finds the session linked to a PR; images can be pasted or referenced by path; Claude reads files fresh on each tool call.
Experimental teams use pending, in-progress and completed tasks, dependencies, explicit assignment and file-locked self-claims. Task directories remain after sessions end.
Subagents, agent view (research preview), agent teams (experimental), dynamic workflows, and Projects. In every approach the workers are Claude sessions; other tools are exposed via MCP. Worktrees isolate parallel sessions.
Kind
documented
Locator
Five ways to work on several tasks; Choose an approach; Check on running work
Subagents have their own context, tools, permissions, model, effort, maxTurns and optional worktree isolation; background subagents run concurrently; results return to the spawning conversation.
Kind
documented
Locator
Built-in subagents; Supported frontmatter fields; Choose a model; Run subagents in foreground or background
Public beta on Team and Enterprise (off by default); runners on customer infrastructure execute cloud sessions; unavailable with ZDR or HIPAA; model inference still via Anthropic API unless Bedrock/Agent Platform configured; session content goes to Anthropic.
Kind
documented
Locator
How self-hosted environments work; Availability and limitations
Subagents have their own context, tools, permissions, model, effort, maxTurns and optional worktree isolation; background subagents run concurrently; results return to the spawning conversation.
Kind
documented
Locator
Built-in subagents; Supported frontmatter fields; Choose a model; Run subagents in foreground or background
Claude can list and message other sessions on the same or another machine or in the cloud; v2.1.224+ on macOS/Linux, v2.1.234+ on Windows; messages are text only.
A JavaScript script Claude writes orchestrates many subagents in the background; on all paid plans, API, Bedrock, Agent Platform and Foundry; /workflows lists runs.
/loop and cron tools are session-scoped (1 minute minimum, recurring tasks expire after 7 days, only fire while Claude Code is running and idle, no catch-up for missed fires); cloud routines minimum 1 hour; Desktop tasks need machine on.
Consumer plans: training on opt-in setting, retention 5 years if allowed else 30 days; commercial plans no training by default, 30-day standard retention, ZDR for qualified Enterprise accounts; local transcripts kept 30 days by default; telemetry/error reporting defaults per provider with env-var opt-outs.
Kind
documented
Locator
Data training policy; Data retention; Telemetry services; Default behaviors by API provider
Lists built-in tools including TaskCreate/TaskUpdate/TaskList, TaskStop, WebFetch, WebSearch, Workflow; task tools are provided by default only on some model families.
Kind
documented
Locator
Tool list; Task tool availability; WebFetch and WebSearch
Projects public beta gradually rolls out to Pro/Max/Ultra; not Team/Enterprise. Parallel cloud threads, optionally local through Remote Control. Projects are single-user and not shareable.
OS-enforced sandbox for shell commands on macOS, Linux and WSL2 (off by default; native Windows runs unsandboxed); built on open source sandbox-runtime; reads allowed broadly by default, writes limited to working directory, network via domain-allowlist proxy; file tools, MCP servers and hooks run outside it.
Kind
documented
Locator
What the sandbox restricts; What runs outside the sandbox
The --worktree or -w flag starts a session in a separate Git working directory, by default .claude/worktrees/<name>, on a new worktree-<name> branch. Subagents can use isolation: worktree. Worktrees share repository history and Git metadata; this documents file separation, not a security sandbox. Source check only; no session was run.
Kind
documented
Locator
Start Claude in a worktree; Isolate subagents with worktrees; What worktrees share with the main checkout
Automatic snapshots before each prompt (100 most recent kept), /rewind restores code and/or conversation; file snapshots swept by retention (~30 days by default).
Newest entry is v2.1.295 (2026-10-08), then v2.1.294 (2026-10-08); v2.1.287 to v2.1.295 are dated Oct 1 to 8. v2.1.295 negotiates MCP protocol 2026-07-28 by default; MCP_PROTOCOL_NEGOTIATION=legacy opts out.
Kind
documented
Locator
Version 2.1.294 (October 8, 2026) and recent entries
/usage shows client-side session cost estimate at list price, plan usage bars and attribution to skills/subagents/plugins/MCP; usage credits past plan limits; --max-budget-usd and --max-turns in print mode; org spend limits for Team/Enterprise; average enterprise cost about $13 per developer per active day.
Kind
documented
Locator
Track your costs; Plan usage breakdown; Usage credits; Manage costs for your organization
Team/Enterprise dashboard at claude.ai/analytics/claude-code with usage and contribution metrics, leaderboard and CSV export; Console dashboard for API customers.
Team/Enterprise dashboard at claude.ai/analytics/claude-code with usage and contribution metrics, leaderboard and CSV export; Console dashboard for API customers.
A JavaScript script Claude writes orchestrates many subagents in the background; on all paid plans, API, Bedrock, Agent Platform and Foundry; /workflows lists runs.
Pro/Max usage limits are shared across Claude and Claude Code; options at limit include usage credits, Console API credits or waiting; monthly API credits included with Max cannot be used in Claude Code; ANTHROPIC_API_KEY overrides subscription auth and causes API charges.
Workflows run up to 16 concurrent agents by default and fewer on hosts or containers with fewer CPUs; CLAUDE_CODE_WORKFLOW_MAX_CONCURRENT_AGENTS accepts 1 to 256 and needs v2.1.269 or later.
Artifacts publish live interactive pages from a session to a private claude.ai URL that can be shared with the organization or a public link; Pro, Max, Team and Enterprise with /login.
Free $0 (Claude Code: No); Pro $17/month annual or $20 monthly (Claude Code: Yes); Max from $100/month (5x or 20x Pro usage); Team standard seat $20 annual/$25 monthly, premium seat $100 annual/$125 monthly (Claude Code included); Enterprise US$20/seat/month plus usage at API rates; usage limits apply; prices exclude tax; model training opt-out on consumer plans, none by default on Team/Enterprise.
Kind
documented
Locator
Individual and Team and Enterprise plan cards; feature comparison
Allow, ask and deny rules for tools and Bash command patterns (for example git commit and git push patterns) layered on permission modes; deny rules apply in every mode; managed settings take precedence over local settings.
Managed GitHub PR Code Review is available to eligible Team/Enterprise organizations and excludes ZDR/HIPAA and customer-managed encryption keys. Inline findings and a non-failing check do not approve or block merges. The current page does not label the service research preview. Local /code-review is a separate route.
Research preview managed GitHub PR review for Team/Enterprise (not with ZDR/HIPAA); findings are inline comments and do not approve or block PRs; /code-review reviews a local diff on other plans.
Releases page marks v2.1.295 (08 Oct 19:48) as Latest, above stored v2.1.294. Issues tab shows 13,889 open (stored 14,591). Main page shows 149.8k stars; no license name or head commit visible.
Kind
documented
Locator
Repository metadata, LICENSE.md, releases and issues
Use is governed by Commercial Terms (Team, Enterprise, API) or Consumer Terms (Free, Pro, Max); products embedding Claude Code must not modify the binary or resell usage on end users behalf.
Kind
documented
Locator
Legal agreements; Can customers offer Claude Code in their products?
CLI, SDK, IDE extensions, subagents, hooks, skills, memory, plugins, MCP, checkpoints, sandboxing, workflows work on every provider; cloud sessions, mobile, Slack, Desktop need a claude.ai subscription.
Kind
documented
Locator
Availability by model provider; Features that require a Claude subscription
The CLI exposes the same agent loop, tools and context management as Claude Code through claude -p for scripts and CI/CD. The installed local core runs with model-provider credentials; this is more than a remote execution worker for the hosted web product.
Kind
documented
Locator
Agent SDK introduction; CLI scripts and CI/CD; Basic usage
Current data policy states opt-in consumer training and five-year versus thirty-day retention by choice; commercial no training by default and thirty-day standard retention. Local plaintext CLI transcripts default to thirty days, with Desktop/Cowork exemptions. Metrics and error reports have separate opt-outs. Error reports default on for direct Claude API Pro, Max and Ultra sign-ins on v2.1.198+, otherwise off. Provider-managed-host metrics are an explicit exception to third-party-provider default-off behavior.
Kind
documented
Locator
Data training policy; Data retention; Telemetry services; Default behaviors by API provider
API prices per million tokens: Fable 5.2 $8 input / $40 output; Opus 5.5 $4 / $20; Sonnet 5.5 $2 / $10; Haiku 5.5 $0.10 / $0.50 for prompts up to 100K tokens and $0.50 / $2.50 above, with separate prompt-caching rates. Batch processing saves 50%; US-only inference costs 1.1x; fast mode for Opus 5.5 and Fable 5.2 costs 2x. Only the Team/Enterprise plan comparison has a Trials row: No for Team and self-serve Enterprise, Yes for sales-assisted Enterprise. The Free, Pro, Max and Ultra cards and comparison mention no trial.
Kind
documented
Locator
Team/Enterprise plan comparison: Trials row; Free/Pro/Max/Ultra plan cards and comparison (no Trials row); Latest models API price cards
References Anthropic Trust Center for SOC 2 Type 2 report and ISO 27001 certificate; documents sandbox, permissions, cloud VM isolation; HackerOne vulnerability reporting.
Kind
documented
Locator
Security foundation; Cloud execution security; Report a vulnerability
Public Claude Code repository and issue navigation reread November 7. It is a report/release repository, not proof of a product source license or reliability.
Kind
documented
Locator
Repository metadata, LICENSE.md, releases and issues
Vendor status page with current component states, incident history and subscription feeds reread November 7. No uptime or response commitment inferred.
Status summary read November 7 reports Claude Code and Console operational at retrieval. This is a point-in-time vendor status report, not uptime evidence.
Kind
documented
Locator
components: Claude Code; Claude Console (platform.claude.com)
References Anthropic Trust Center for SOC 2 Type 2 report and ISO 27001 certificate; documents sandbox, permissions, cloud VM isolation; HackerOne vulnerability reporting.
Kind
documented
Locator
Security foundation; Cloud execution security; Report a vulnerability