Capture
One read of Claude Code CLI
4 pages moved out of 191 read.
context-window Changed · +34 / -20 lines
from line 87
tokens: 2400, color: '#8A8880', vis: 'brief', + restoredAfterCompact: true, desc: 'Main auth file. You see "Read auth.ts" in your terminal, but the 2,400 tokens of file content only Claude sees.', tip: 'File reads dominate context usage. Be specific in prompts ("fix the bug in auth.ts") so Claude reads fewer files. For research-heavy tasks, use a subagent.', link: null
from line 98
tokens: 1100, color: '#8A8880', vis: 'brief', + restoredAfterCompact: true, desc: 'Following imports to the token module. Shown as a one-liner in your terminal.', link: null }, {
from line 108
tokens: 380, color: '#4A9B8E', vis: 'brief', + restoredAfterCompact: true, desc: 'This rule in `.claude/rules/` has a `paths:` pattern matching `src/api/**`. It loaded automatically when Claude read a file in that directory. You see "Loaded .claude/rules/api-conventions.md" in your terminal, but not the rule content.', link: '/en/memory#path-specific-rules' }, {
from line 118
tokens: 1800, color: '#8A8880', vis: 'brief', + restoredAfterCompact: true, desc: 'Tracing the auth flow deeper.', link: null }, {
from line 128
tokens: 1600, color: '#8A8880', vis: 'brief', + restoredAfterCompact: true, desc: 'Checking existing tests for expected behavior.', link: null }, {
from line 138
tokens: 290, color: '#4A9B8E', vis: 'brief', + restoredAfterCompact: true, desc: 'Another path-scoped rule, this one matching `*.test.ts` files. Triggered when Claude read auth.test.ts. Shown as a one-line "Loaded" notice.', link: '/en/memory#path-specific-rules' }, {
from line 337
tokens: 620, color: '#558A42', vis: 'brief', - desc: 'You invoked a skill that has `disable-model-invocation: true`. Its description was not in the skill index at startup, so it cost zero context until this moment. Now the full skill content loads and Claude follows its instructions to stage, commit, and push your changes.', + restoredAfterCompact: true, + desc: 'You invoked a skill that has `disable-model-invocation: true`. Its description was not in the skill index at startup, so it cost zero context until this moment. Now the full skill content loads and Claude follows its instructions to stage, commit, and push your changes. After `/compact`, Claude Code re-injects the body of each skill you invoked, capped at 5,000 tokens per skill.', tip: 'Set `disable-model-invocation: true` on skills with side effects like committing, deploying, or sending messages. They stay out of context entirely until you need them.', link: '/en/skills#control-who-invokes-a-skill' }, {}, {
from line 534
} {} const autoLoads = nonCompact.filter(e => e.kind === 'auto' && e.t < STARTUP_END && !e.noSurviveCompact); + const restored = nonCompact.filter(e => e.restoredAfterCompact); const summarized = nonCompact.filter(e => e.t >= STARTUP_END && e.kind !== 'sub'); const sumTokens = summarized.reduce((s, e) => s + e.tokens, 0); const summaryBlock = {
from line 544
tokens: Math.round(sumTokens * 0.12), color: '#A09E96', vis: 'hidden', - desc: `All ${summarized.length} conversation events condensed into one structured summary. The summary keeps: your requests and intent, key technical concepts, files examined or modified with important code snippets, errors and how they were fixed, pending tasks, and current work. It replaces the verbatim conversation: full tool outputs and intermediate reasoning are gone. Claude can still reference the work but won't have the exact code it read earlier.`, + desc: `All ${summarized.length} conversation events condensed into one structured summary. The summary keeps: your requests and intent, key technical concepts, files examined or modified with important code snippets, errors and how they were fixed, pending tasks, and current work. It replaces the verbatim conversation: full tool outputs and intermediate reasoning are gone. Claude Code re-reads the files modified most recently and re-injects the skills you invoked, listed below. Claude can still reference the rest of the work but won't have the exact content.`, link: '/en/how-claude-code-works#the-context-window' }; return { - visible: [...autoLoads, summaryBlock], + visible: [...autoLoads, summaryBlock, ...restored], preCompactTotal: nonCompact.reduce((s, e) => s + e.tokens, 0) }; }, [preCompactVisible, isCompacted]);
from line 621
if (detailRef.current) detailRef.current.scrollTop = 0; }, [hovEvent]); const focusT = hovEvent ? hovEvent.t : time; - const takeaway = isCompacted ? 'Compaction replaces the conversation with a structured summary. System prompt, CLAUDE.md, memory, and MCP tools reload automatically. The skill listing is the one exception. Only skills you actually invoked are preserved.' : focusT < STARTUP_END ? 'A lot loads before you type anything. CLAUDE.md, memory, skills, and MCP tools are all in context before your first prompt.' : focusT < 0.28 ? "Your prompt is tiny compared to what's already loaded. Most of Claude's context is project knowledge, not your words." : focusT < 0.50 ? 'Each file Claude reads grows the context. Path-scoped rules load automatically alongside matching files.' : focusT < 0.71 ? 'Hooks fire automatically on tool events. Output reaches Claude via additionalContext JSON. Exit code 2 surfaces stderr to Claude. Plain stdout on exit 0 goes to the debug log, not the transcript.' : focusT < 0.79 ? 'Follow-up questions keep building on the same context. Everything from earlier is still there.' : focusT < 0.87 ? "The subagent works in its own separate context window. None of its file reads touch yours. Only the final summary comes back." : focusT < 0.88 ? 'Bang commands run in your shell and prefix the output to your next message. Useful for grounding Claude in command results without it running them.' : focusT < 0.90 ? 'User-only skills stay out of context entirely until you invoke them. The skill index at startup only lists skills Claude can call on its own.' : '/compact summarizes the conversation to free space while keeping key information. In a real session, run it when context starts affecting performance or before a long new task.'; - const terminalView = isCompacted ? 'A "Conversation compacted" message. The summarization happens silently.' : focusT < STARTUP_END ? 'The input box, waiting for your first message. Everything above loads silently before you type anything.' : focusT < 0.28 ? 'Your prompt. Claude hasn\'t started working yet.' : focusT < 0.52 ? 'Your prompt and "Reading files...". Rules show as one-line "Loaded" notices, not their content.' : focusT < 0.72 ? "Claude's response and file diffs. Hooks fire silently. Tool output like npm test shows as a brief summary, not the full content." : focusT < 0.79 ? 'Your follow-up prompt.' : focusT < 0.86 ? "A brief notice that a subagent is working, then its result. You don't see the subagent's individual file reads." : focusT < 0.90 ? "Claude's response, your git status output, and the commit-push skill running." : 'Your full conversation. /compact is available to run.'; + const takeaway = isCompacted ? 'Compaction replaces the conversation with a structured summary. System prompt, CLAUDE.md, memory, and MCP tools reload automatically. Claude Code also re-reads up to five of the files modified most recently, reloads the rules that match them, and re-injects the skills you invoked. The skill listing does not reload.' : focusT < STARTUP_END ? 'A lot loads before you type anything. CLAUDE.md, memory, skills, and MCP tools are all in context before your first prompt.' : focusT < 0.28 ? "Your prompt is tiny compared to what's already loaded. Most of Claude's context is project knowledge, not your words." : focusT < 0.50 ? 'Each file Claude reads grows the context. Path-scoped rules load automatically alongside matching files.' : focusT < 0.71 ? 'Hooks fire automatically on tool events. Output reaches Claude via additionalContext JSON. Exit code 2 surfaces stderr to Claude. Plain stdout on exit 0 goes to the debug log, not the transcript.' : focusT < 0.79 ? 'Follow-up questions keep building on the same context. Everything from earlier is still there.' : focusT < 0.87 ? "The subagent works in its own separate context window. None of its file reads touch yours. Only the final summary comes back." : focusT < 0.88 ? 'Bang commands run in your shell and prefix the output to your next message. Useful for grounding Claude in command results without it running them.' : focusT < 0.90 ? 'User-only skills stay out of context entirely until you invoke them. The skill index at startup only lists skills Claude can call on its own.' : '/compact summarizes the conversation to free space while keeping key information. In a real session, run it when context starts affecting performance or before a long new task.'; + const terminalView = isCompacted ? 'A "Conversation compacted" message, then a one-line "Read auth.ts" for each re-read file and "Skills restored (commit-push)". The rules show as "Loaded" lines on Claude\'s next turn. None of the content itself appears.' : focusT < STARTUP_END ? 'The input box, waiting for your first message. Everything above loads silently before you type anything.' : focusT < 0.28 ? 'Your prompt. Claude hasn\'t started working yet.' : focusT < 0.52 ? 'Your prompt and "Reading files...". Rules show as one-line "Loaded" notices, not their content.' : focusT < 0.72 ? "Claude's response and file diffs. Hooks fire silently. Tool output like npm test shows as a brief summary, not the full content." : focusT < 0.79 ? 'Your follow-up prompt.' : focusT < 0.86 ? "A brief notice that a subagent is working, then its result. You don't see the subagent's individual file reads." : focusT < 0.90 ? "Claude's response, your git status output, and the commit-push skill running." : 'Your full conversation. /compact is available to run.'; const mono = 'var(--font-mono, ui-monospace, SFMono-Regular, Menlo, monospace)'; const renderWithCode = s => s.split('`').map((part, i) => i % 2 === 1 ? <code key={i} style={{ fontFamily: mono,
from line 948
lineHeight: 1.5, marginTop: 4 }}> - This is what's left in context: startup content, which lives outside the message history and reloads after compaction, plus a structured summary of the entire conversation. Skill descriptions don't reload. + This is what's left in context: startup content, which lives outside the message history and reloads after compaction, a structured summary of the entire conversation, the files modified most recently, which Claude Code re-reads along with the rules that match them, and the body of each skill you invoked. Skill descriptions don't reload. </div> </div>} {time > 0 && visible.length > 0 && <div style={{
from line 971
const enteringSubagent = isSub && prevKind !== 'sub'; const leavingSubagent = prevKind === 'sub' && !isSub; let showPhase = null; - if (evt.kind === 'user' && prevKind !== 'user') showPhase = 'You'; else if (evt.kind === 'claude' && prevKind === 'user') showPhase = 'Claude works'; else if (evt.label === 'Conversation summary') showPhase = 'Summarized by /compact'; + if (isCompacted && evt.restoredAfterCompact) { + if (prevKind === 'compact') showPhase = 'Restored after /compact'; + } else if (evt.kind === 'user' && prevKind !== 'user') showPhase = 'You'; else if (evt.kind === 'claude' && prevKind === 'user') showPhase = 'Claude works'; else if (evt.label === 'Conversation summary') showPhase = 'Summarized by /compact'; const isNewRow = isCompacted && !(evt.kind === 'auto' && evt.t < STARTUP_END); return <div key={evt.label + evt.t} className={isNewRow ? 'cw-compacted-row' : ''} style={isNewRow ? { animationDelay: `${i * 60}ms`
from line 1589
## What survives compaction -When a long session compacts, Claude Code summarizes the conversation history to fit the context window. As of v2.1.198, the summarization request inherits your session's [extended thinking](/docs/en/model-config#extended-thinking) configuration, so it reasons with thinking enabled when your session has it enabled and stays off otherwise. Thinking affects only how the summary is produced; your session settings are unchanged afterward. What happens to your instructions depends on how they were loaded: +When a long session compacts, Claude Code summarizes the conversation history to fit the context window. As of v2.1.198, the summarization request inherits your session's [extended thinking](/docs/en/model-config#extended-thinking) configuration, so it reasons with thinking enabled when your session has it enabled and stays off otherwise. Thinking affects only how the summary is produced; your session settings are unchanged afterward. What happens to each kind of content depends on how it was loaded: -| Mechanism | After compaction | -| :---------------------------------------- | :------------------------------------------------------------------------------------------ | -| System prompt and output style | Unchanged; not part of message history | -| Project-root CLAUDE.md and unscoped rules | Re-injected from disk | -| Auto memory | Re-injected from disk | -| Rules with `paths:` frontmatter | Lost until a matching file is read again | -| Nested CLAUDE.md in subdirectories | Lost until a file in that subdirectory is read again | -| Invoked skill bodies | Re-injected, capped at 5,000 tokens per skill and 25,000 tokens total; oldest dropped first | -| Hooks | Not applicable; hooks run as code, not context | +| Mechanism | After compaction | +| :------------------------------------------------------------------------------------------------------- | :------------------------------------------------------------------------------------------ | +| System prompt and output style | Unchanged; not part of message history | +| Project-root CLAUDE.md and unscoped rules | Re-injected from disk | +| Auto memory | Re-injected from disk | +| The plan Claude wrote in [plan mode](/docs/en/permission-modes#analyze-before-you-edit-with-plan-mode) | Re-injected from disk | +| Rules with `paths:` frontmatter | Claude Code reloads them as Claude reads files they match | +| Nested CLAUDE.md in subdirectories | Claude Code reloads them as Claude reads files in that subdirectory | +| Files Claude read or edited | Claude Code re-reads up to five, most recently modified first | +| Invoked skill bodies | Re-injected, capped at 5,000 tokens per skill and 25,000 tokens total; oldest dropped first | +| Context that hooks added earlier | Summarized with the rest of the conversation | +| [SessionStart hooks](/docs/en/hooks-guide#re-inject-context-after-compaction) that match the `compact` source | Claude Code runs them and adds their output to the compacted context | -Path-scoped rules and nested CLAUDE.md files load into message history when their trigger file is read, so compaction summarizes them away with everything else. They reload the next time Claude reads a matching file. If a rule must persist across compaction, drop the `paths:` frontmatter or move it to the project-root CLAUDE.md. +Path-scoped rules and nested CLAUDE.md files load into message history when their trigger file is read, so compaction summarizes them away with everything else. Right after compaction, Claude Code re-reads up to five of the files Claude has read or edited in the session, choosing the ones modified most recently, and reloads the rules and nested CLAUDE.md files that apply to those files. A file over 5,000 tokens comes back as a path reference without its content, shown as `Referenced file` instead of `Read`. Its rules still reload. If a rule must persist across compaction, drop the `paths:` frontmatter or move it to the project-root CLAUDE.md. Skill bodies are re-injected after compaction, but large skills are truncated to fit the per-skill cap, and the oldest invoked skills are dropped once the total budget is exceeded. Truncation keeps the start of the file, so put the most important instructions near the top of `SKILL.md`.
from line 1615
You can also act before the automatic pass runs: * **Compact with a focus**: run `/compact` with instructions, like `/compact focus on the auth bug fix`, before starting a long new task. The summary keeps what you choose instead of what the automatic pass guesses is important. +* **Compact part of the conversation**: run `/rewind`, select a message, and choose **Summarize from here** or **Summarize up to here**. See [Rewind and summarize](/docs/en/checkpointing#rewind-and-summarize) for what each option keeps and how to guide the summary. * **Compact earlier**: run [`/autocompact`](/docs/en/commands#all-commands) with a token count, like `/autocompact 500k`, to set how full the context window gets before the automatic pass runs. See [Set the auto-compact window](/docs/en/model-config#set-the-auto-compact-window) for accepted values and overrides. * **Clear between tasks**: run `/clear` when switching to unrelated work. Old conversation crowds out the files you need next and costs tokens on every message. * **Delegate large reads**: send research to a [subagent](/docs/en/sub-agents) so the file contents stay in its context window, not yours.
managed-mcp Changed · +10 / -10 lines
from line 126
Allowlists and denylists filter which configured servers are allowed to load. They aren't a registry: a server still has to be added by a user, a plugin, or `managed-mcp.json` before the allowlist or denylist applies to it. To deploy servers to users, use [`managed-mcp.json`](#exclusive-control-with-managed-mcp-json). Both lists also filter servers passed with the [`--mcp-config` CLI flag](/docs/en/cli-reference#cli-flags); `--strict-mcp-config` limits which configuration files load and doesn't bypass either list. -To make the allowlist authoritative, set `allowedMcpServers` and `allowManagedMcpServersOnly: true` together in a [managed settings source](/docs/en/admin-setup#decide-how-settings-reach-devices), such as server-managed settings or a deployed `managed-settings.json` file. [Restrict the allowlist to managed settings only](#restrict-the-allowlist-to-managed-settings-only) shows the configuration. Without `allowManagedMcpServersOnly`, allowlists from every settings source merge, including a user's own `~/.claude/settings.json`, so a user can broaden what your allowlist permits. Denylists merge from every source regardless. +To make the allowlist authoritative, set `allowedMcpServers` and `allowManagedMcpServersOnly: true` together in a [managed settings source](/docs/en/admin-setup#decide-how-settings-reach-devices), such as server-managed settings or a deployed `managed-settings.json` file. [Restrict the allowlist to managed settings only](#restrict-the-allowlist-to-managed-settings-only) shows the configuration. Without `allowManagedMcpServersOnly`, allowlists from every settings scope merge, including a user's own `~/.claude/settings.json`, so a user can broaden what your allowlist permits. Denylists merge from every scope regardless. <Note> `allowManagedMcpServersOnly` is separate from `allowManagedPermissionRulesOnly`, which locks down [permission rules](/docs/en/permissions#managed-settings) only. Setting that flag does not enforce the MCP allowlist.
from line 166
Before loading a server, including one from `managed-mcp.json`, Claude Code runs three checks in order: -1. **Merge the lists.** Allowlist and denylist entries from every settings source combine into one allowlist and one denylist. When `allowManagedMcpServersOnly` is `true`, only the managed allowlist is kept; the denylist always merges from every source. +1. **Merge the lists.** Allowlist and denylist entries from every settings scope combine into one allowlist and one denylist, with the managed scope's lists read from the one [managed source Claude Code selects](/docs/en/managed-settings#precedence-within-the-managed-tier). When `allowManagedMcpServersOnly` is `true`, only the managed allowlist is kept; the denylist always merges from every scope. 2. **Check the denylist.** A server that matches any denylist entry, by URL, command, or name, is blocked. Nothing overrides a denylist match. 3. **Check the allowlist.** If `allowedMcpServers` isn't set anywhere, every server that passed the denylist loads. If it is set, what the server must match depends on its type, shown in the table below.
from line 334
} ``` -When `allowManagedMcpServersOnly` is `true`, allowlists from user, project, and local settings are ignored. The denylist still merges from all sources, so users can always block servers for themselves. +When `allowManagedMcpServersOnly` is `true`, allowlists from user, project, and local settings are ignored. The denylist still merges from every settings scope, so users can always block servers for themselves. ## How restrictions appear to users
from line 357
Every file and setting this page covers, what it controls, and how to deliver it: -| Surface | What it controls | Where it lives | How to deliver | -| :--------------------------- | :------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | :-------------------------------------------------------------------------------------------------------------------------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -| `managed-mcp.json` | Fixed server set, exclusive control | System path: `/Library/Application Support/ClaudeCode/`, `/etc/claude-code/`, or `C:\Program Files\ClaudeCode\` | MDM, GPO, fleet management, or any process with administrator privileges. Cannot be set through server-managed settings | -| `allowedMcpServers` | Allowlist of permitted servers | Any [settings file](/docs/en/settings#where-settings-live); entries from every source merge unless `allowManagedMcpServersOnly` is set | For enforcement, a [managed settings source](/docs/en/admin-setup#decide-how-settings-reach-devices): server-managed settings, `managed-settings.json`, MDM profile, or registry | -| `deniedMcpServers` | Denylist of blocked servers | Any settings file; entries from every source merge | Same as `allowedMcpServers` | -| `allowManagedMcpServersOnly` | Locks the allowlist to managed sources only | Managed settings sources only; the setting has no effect elsewhere | Same as `allowedMcpServers` | -| `allowAllClaudeAiMcps` | Loads the claude.ai connectors Claude Code fetches itself alongside `managed-mcp.json`. [Connectors delivered to cloud sessions stay suppressed](#allow-claude-ai-connectors-alongside-the-managed-set) | Managed settings sources only; the setting has no effect elsewhere | Same as `allowedMcpServers` | +| Surface | What it controls | Where it lives | How to deliver | +| :--------------------------- | :------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | :------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | +| `managed-mcp.json` | Fixed server set, exclusive control | System path: `/Library/Application Support/ClaudeCode/`, `/etc/claude-code/`, or `C:\Program Files\ClaudeCode\` | MDM, GPO, fleet management, or any process with administrator privileges. Cannot be set through server-managed settings | +| `allowedMcpServers` | Allowlist of permitted servers | Any [settings scope](/docs/en/settings#where-settings-live); Claude Code merges the lists from every scope unless `allowManagedMcpServersOnly` is set, and takes the managed scope's list from the one [managed source it selects](/docs/en/managed-settings#precedence-within-the-managed-tier) | For enforcement, a [managed settings source](/docs/en/admin-setup#decide-how-settings-reach-devices): server-managed settings, `managed-settings.json`, MDM profile, or registry | +| `deniedMcpServers` | Denylist of blocked servers | Any settings scope; Claude Code merges the lists from every scope and takes the managed scope's list from the one managed source it selects | Same as `allowedMcpServers` | +| `allowManagedMcpServersOnly` | Locks the allowlist to managed sources only | Managed settings sources only; the setting has no effect elsewhere | Same as `allowedMcpServers` | +| `allowAllClaudeAiMcps` | Loads the claude.ai connectors Claude Code fetches itself alongside `managed-mcp.json`. [Connectors delivered to cloud sessions stay suppressed](#allow-claude-ai-connectors-alongside-the-managed-set) | Managed settings sources only; the setting has no effect elsewhere | Same as `allowedMcpServers` | ## Related resources
quickstart Changed · +5 / -5 lines
from line 25
<Tab title="Native Install (Recommended)"> **macOS, Linux, WSL:** - ```bash theme={null} theme={null} theme={null} theme={null} theme={null} + ```bash theme={null} curl -fsSL https://claude.ai/install.sh | bash ``` **Windows PowerShell:** - ```powershell theme={null} theme={null} theme={null} theme={null} theme={null} + ```powershell theme={null} irm https://claude.ai/install.ps1 | iex ``` **Windows CMD:** - ```batch theme={null} theme={null} theme={null} theme={null} theme={null} + ```batch theme={null} curl -fsSL https://claude.ai/install.cmd -o install.cmd && install.cmd && del install.cmd ```
from line 53
</Tab> <Tab title="Homebrew"> - ```bash theme={null} theme={null} theme={null} theme={null} theme={null} + ```bash theme={null} brew install --cask claude-code ```
from line 65
</Tab> <Tab title="WinGet"> - ```powershell theme={null} theme={null} theme={null} theme={null} theme={null} + ```powershell theme={null} winget install Anthropic.ClaudeCode ```
settings-reference Changed · +3 / -3 lines
This page is larger than the 256 KiB this site keeps, so one side of the diff below stops where the stored text does.
from line 1318
} ``` -Permission rules layer on top of every mode: `deny` rules block in every mode, including `bypassPermissions`. See [Permission modes](/docs/en/permission-modes). `manual` names the permission mode labeled Manual in the CLI and the VS Code extension; the alias requires Claude Code v2.1.200 or later. Before v2.1.142, project settings could set `auto`. In Claude Code on the web, Claude Code honors only `acceptEdits`, `plan`, `default`, and `auto` from this key. For conversations the VS Code extension starts, see [which setting the extension reads for the starting permission mode](/docs/en/permission-modes#switch-permission-modes). +Permission rules layer on top of every mode: `deny` rules block in every mode, including `bypassPermissions`. See [Permission modes](/docs/en/permission-modes). `manual` names the permission mode labeled Manual in the CLI and the VS Code extension; the alias requires Claude Code v2.1.200 or later. In Claude Code on the web, Claude Code honors only `acceptEdits`, `plan`, `default`, and `auto` from this key. For conversations the VS Code extension starts, see [which setting the extension reads for the starting permission mode](/docs/en/permission-modes#switch-permission-modes). ### `permissions.disableBypassPermissionsMode`
from line 3683
### `blockedMarketplaces` -Block plugin marketplace sources for your organization. Claude Code checks t +Block plugin marketplace sources for your organization. Claude Code checks the blocklist on marketplace add and on plugin instal