# Crewship

- [Crewship Documentation](https://docs.crewship.ai/index.md): Self-hosted runtime for AI coding agents. Per-crew Linux containers, runs entirely on your hardware, full backup by default — code, data, conversations, cost ledger.
- [Quick Start](https://docs.crewship.ai/quickstart.md): Get Crewship running on your hardware in 60 seconds, then create your first crew in under 5 minutes.
- [Find your way around Crewship](https://docs.crewship.ai/guides/task-index.md): Choose a starting page by the job you want to accomplish instead of by the command or API resource name.
- [First projects](https://docs.crewship.ai/guides/first-projects.md): Build a small Crewship workflow in five runnable rungs: create a crew, run an agent, share memory, schedule a routine, and inspect the journal.
- [Crewship concepts and glossary](https://docs.crewship.ai/concepts.md): Understand which Crewship object owns work, memory, credentials, approvals, cost, and scheduling before choosing a command or manifest kind.
- [Crewship use cases](https://docs.crewship.ai/guides/use-cases.md): Match common engineering operations to a Crewship crew, routine, credential policy, and the command or manifest that starts them.
- [Writing a good task for an agent](https://docs.crewship.ai/guides/writing-good-tasks.md): Turn a product goal into an agent task with the right repository context, acceptance checks, boundaries, and durable output.
- [When not to use Crewship](https://docs.crewship.ai/guides/when-not-to-use.md): Recognize the jobs Crewship is not designed for, including low-latency services, unreviewed production changes, and work with missing context or credentials.
- [Philosophy](https://docs.crewship.ai/philosophy.md): Why Crewship exists and what it stands for. The manifesto behind the runtime — written for builders who want to take agents to production.
- [Production Agent Checklist](https://docs.crewship.ai/production-checklist.md): Six things a production agent system needs — and how Crewship covers all of them. A neutral framework for evaluating any agent stack against real operational requirements.
- [Architecture](https://docs.crewship.ai/architecture.md): How Crewship runs AI coding agents — one Linux container per crew, sidecar proxy, Unix socket IPC, and a single-binary deployment model that runs entirely on your hardware.
- [Install](https://docs.crewship.ai/guides/install.md): Supported ways to install Crewship: Homebrew tap, signed curl | bash installer, Debian/RPM packages, or Docker.
- [Developer Installation](https://docs.crewship.ai/guides/installation.md): Build Crewship from source, run the Go + Next.js dev servers via dev.sh, and lay out multi-instance and remote-dev environments.
- [Onboarding](https://docs.crewship.ai/guides/onboarding.md): First-run flow from empty database to a logged-in user with a primary LLM provider configured and a first crew ready to run.
- [Authentication](https://docs.crewship.ai/guides/auth.md): Sign-in, signup, password recovery, OAuth, sessions, and the CLI tokens / device-pairing flow — every way a user proves identity to Crewship.
- [Admin CLI](https://docs.crewship.ai/guides/admin-cli.md): Host-side `crewship admin` subcommands that act on the SQLite database file on this machine — the recovery surface for an operator who's locked out of the UI. Says --local, because the CLI cannot guess which database you mean.
- [Your First Crew](https://docs.crewship.ai/guides/first-crew.md): Create a team of AI agents, assign roles and credentials, and run your first mission.
- [Chat & Sessions](https://docs.crewship.ai/guides/chat-sessions.md): Full-page chat with agent sessions, direct messages, groups and channels, attachments with previews, message feedback, and the Artifacts / Work side panels.
- [Chat Surface Limits](https://docs.crewship.ai/guides/chat-surface-limits.md): Every bound the chat surface applies — the value, what happens when you reach it, and whether anything tells you.
- [Ask forms](https://docs.crewship.ai/guides/ask-forms.md): Chips that open a short questionnaire, render the answers into an ordinary message, and record what was answered alongside it.
- [Chat demo walkthrough](https://docs.crewship.ai/guides/chat-demo-walkthrough.md): Ten minutes from a seeded install to having seen every part of the chat surface work — and four of them refuse, on purpose.
- [Demo use cases](https://docs.crewship.ai/guides/demo-use-cases.md): Twelve product use cases as twelve scripts — run one at a time against a seeded install, each narrated for an audience and verified by assertions.
- [Watching Agent Runs](https://docs.crewship.ai/guides/watching-agent-runs.md): Follow an agent run from a shell over plain HTTP — newline-delimited JSON, resumable, no WebSocket client required.
- [Workspace Manifests](https://docs.crewship.ai/guides/manifests.md): Declare crews, agents, skills, credentials and sidecar services as YAML. Apply converges your workspace toward the manifest — like Terraform, for AI agents.
- [Devcontainers & Runtime Images](https://docs.crewship.ai/guides/devcontainers.md): Bring your own Linux base image, declare devcontainer features and mise tools, get a cached per-crew image. Sidecar bind-mounts from the host.
- [Keeping crew images current](https://docs.crewship.ai/guides/crew-image-freshness.md): Crewship checks daily whether a crew's container is running an older build than its image tag now points at, tells you, and gives you one command to fix it.
- [Credentials](https://docs.crewship.ai/guides/credentials.md): Manage encrypted API keys with priority-based selection, round-robin rotation, and Keeper security gating.
- [Auto-managed credentials](https://docs.crewship.ai/guides/auto-managed-credentials.md): How sidecar database passwords are generated and stored automatically; rotation is available as a follow-up action, not at apply time.
- [Model Discovery](https://docs.crewship.ai/guides/model-discovery.md): List the models a provider can serve, live from the provider or from a curated fallback, and validate an agent's model against that set.
- [Multi-provider LLM configuration](https://docs.crewship.ai/guides/multi-provider-llm.md): The provider registry, the two configurable codecs, and the embedded model catalog that prices them.
- [Skills & Marketplace](https://docs.crewship.ai/guides/skills.md): Add capabilities to agents with system prompt extensions -- built-in skills, custom skills, and the SKILL.md format.
- [Author a Skill in 10 Minutes](https://docs.crewship.ai/guides/skills-authoring.md): End-to-end walkthrough: idea → SKILL.md → import → assign → invoke.
- [Agent-Authored Skills](https://docs.crewship.ai/guides/skills-agent-authored.md): How an agent turns a workflow it just performed into a reusable crew skill — staged, reviewed in the inbox, and published.
- [SKILL.md Format Reference](https://docs.crewship.ai/guides/skills-format.md): The frontmatter and body conventions parsed from a SKILL.md file.
- [Orchestration](https://docs.crewship.ai/guides/orchestration.md): Multi-agent missions with task dependencies, retry loops, deadlock detection, and circuit breakers.
- [The issue detail](https://docs.crewship.ai/guides/issue-detail.md): One issue, one screen — the card surface behind /issues/<identifier> and the /issues centre pane, and the URL that now carries your selection.
- [Mentioning an agent in an issue](https://docs.crewship.ai/guides/issue-mentions.md): @mention an agent in an issue comment: the picker, how the mention is stored, what wakes the agent, and the limits that bound it.
- [Inbox](https://docs.crewship.ai/guides/inbox.md): Per-user actionable feed across every crew and workspace — waitpoints, escalations, failed runs, and agent messages in one Linear-Triage list.
- [Routines](https://docs.crewship.ai/guides/routines.md): Declarative AI workflow recipes — AI-authored, repeatable, with cron/webhook/manual triggers, two-tier execution, HITL waitpoints, immutable versions.
- [Routines cookbook](https://docs.crewship.ai/guides/routines-cookbook.md): Worked examples and patterns for production-ready routines: gates, tier escalation, outcomes rubrics, eval-driven promotion
- [Demo packs](https://docs.crewship.ai/guides/demo-packs.md): The legacy GitHub demo packs — what each one watched, what it cost, and how to verify one against a workspace that still carries it.
- [Routines: Migration & Compatibility](https://docs.crewship.ai/guides/routines-migration.md): Naming change Pipeline → Routine, what stays the same, what's new, and how existing automations keep working without changes.
- [Outbound notifications](https://docs.crewship.ai/guides/notifications.md): Email, signed-webhook, chat and push delivery for approvals, escalations, failed runs, issue activity, chat replies, and more — with a per-user preference matrix and anti-storm rate limiting.
- [CLI Adapters](https://docs.crewship.ai/guides/cli-adapters.md): How Crewship invokes vendor coding CLIs (Claude Code, Codex, Gemini, OpenCode, Cursor, Factory Droid) and the reasoning behind which ones are supported.
- [CLI Pairing](https://docs.crewship.ai/guides/cli-pairing.md): Device-code handoff that authenticates a locally-installed `crewship` CLI against a Crewship server using an 8-character one-time code shown in the browser.
- [Claude Code](https://docs.crewship.ai/guides/cli/claude-code.md): Anthropic's coding agent (CLAUDE_CODE adapter)
- [Codex CLI](https://docs.crewship.ai/guides/cli/codex.md): OpenAI's coding agent (CODEX_CLI adapter)
- [Gemini CLI](https://docs.crewship.ai/guides/cli/gemini.md): Google's coding agent (GEMINI_CLI adapter)
- [OpenCode](https://docs.crewship.ai/guides/cli/opencode.md): sst.dev's BYOK multi-provider CLI (OPENCODE adapter)
- [Cursor CLI](https://docs.crewship.ai/guides/cli/cursor.md): Cursor's headless agent (CURSOR_CLI adapter)
- [Factory Droid](https://docs.crewship.ai/guides/cli/factory-droid.md): Factory's autonomous coding agent (FACTORY_DROID adapter)
- [MCP across all 6 CLIs](https://docs.crewship.ai/guides/mcp-multi-cli.md): Side-by-side MCP config schema reference for every Crewship CLI adapter
- [Restricted login identity proof](https://docs.crewship.ai/guides/restricted-login-identity-proof.md)
- [Keeper](https://docs.crewship.ai/guides/keeper.md): Local AI security gatekeeper for credential access control with L1-L4 security levels and prompt injection defense.
- [Autonomy + self-learning](https://docs.crewship.ai/guides/autonomy-and-self-learning.md): How the per-crew autonomy slider, per-crew behavior_mode, and per-agent self_learning_enabled flag compose into the day-to-day governance posture for an agent.
- [Keeper Reviews — admin panel workflow](https://docs.crewship.ai/guides/keeper-reviews-panel.md): Operator playbook for the Keeper Phase 2 reviews panel in the admin UI — triaging the four evaluator queues, investigating decisions, approving pending lesson proposals.
- [Ephemeral agents — hire, ghost, rehire](https://docs.crewship.ai/guides/ephemeral-agents.md): Operator playbook for the short-lived contractor-agent lifecycle: hire with a TTL, watch it ghost when the TTL elapses, rehire if the work isn't done.
- [Integrations](https://docs.crewship.ai/guides/integrations.md): Connect agents to external services via OAuth 2.0, MCP servers, and the official MCP registry.
- [Git links](https://docs.crewship.ai/guides/git-links.md): Attach a GitHub pull request or GitLab merge request to an issue — no webhooks, no publicly reachable instance.
- [Issue attachments](https://docs.crewship.ai/guides/issue-attachments.md): Attach files to an issue — and let the agent working it actually read them.
- [Templates](https://docs.crewship.ai/guides/templates.md): Deploy pre-configured crews with agents from blueprints, and use workflow templates for mission patterns.
- [Memory System](https://docs.crewship.ai/guides/agent-memory.md): 3-tier persistent memory (agent, crew, session) with FTS5 search, dynamic budget allocation, and multi-scope queries.
- [Memory Portability](https://docs.crewship.ai/guides/memory-portability.md): Export agent memory to a readable OKF bundle, and import memory from NanoClaw, OpenClaw or another Crewship.
- [Memory Provider interface](https://docs.crewship.ai/guides/memory-provider.md): How to add a new memory backend (HTTP-backed external store, vector recall service, conversation-history platform) to Crewship without forking the dispatcher.
- [AgentBrief — sub-agent briefing primitive](https://docs.crewship.ai/guides/agent-brief.md): How a LEAD agent hands a curated context slice to a freshly hired sub-agent — the structured replacement for the all-or-nothing SkipConvHistory boolean.
- [Episodic Memory](https://docs.crewship.ai/guides/episodic-memory.md): Vector-similarity recall over the Crew Journal with selective embedding and SQLite BLOB brute-force cosine scan.
- [Conversation Search](https://docs.crewship.ai/guides/conversation-search.md): Keyword (BM25) search across past chat sessions — one agent or the whole workspace — ranked by relevance.
- [Consolidate](https://docs.crewship.ai/guides/consolidate.md): Nightly workers that compact old journal entries and extract semantic rules into crew memory.
- [Memory Observability](https://docs.crewship.ai/guides/memory-observability.md): Operator drill-down into memory_versions, per-workspace retention policy, and HITL diff preview for consolidator proposals.
- [LLM Middleware](https://docs.crewship.ai/guides/llm-middleware.md): The unified LLM call stack: telemetry → paymaster → lookout → raw provider. Composition and layer order rationale.
- [Encryption at rest](https://docs.crewship.ai/guides/encryption-at-rest.md): Operator-side setup for encrypting agent memory directories (PERSONA.md, peer cards, AGENT.md) with LUKS or eCryptfs. GDPR baseline for EU-region installs.
- [Crew Journal](https://docs.crewship.ai/guides/crew-journal.md): Append-only event stream that is the single source of truth for every observable action in the platform.
- [Activity](https://docs.crewship.ai/guides/activity.md): The workspace activity stream — four lenses over the workflows that ran, each row leading to the thing itself: a workflow's causal graph, a run's steps, an issue, an agent.
- [Bottom dock — context-aware inspector](https://docs.crewship.ai/guides/bottom-dock.md): The collapsible panel at the bottom of the Crews page adapts its tabs to the selected agent or crew — terminal, files, messages, logs — so the live, hands-on layer is always one click away. Issues carry their own drawer.
- [Tracing (OpenTelemetry)](https://docs.crewship.ai/guides/tracing.md): OpenTelemetry GenAI spans with W3C Trace Context propagation into journal entries.
- [Otlp setup](https://docs.crewship.ai/observability/otlp-setup.md)
- [Prometheus Metrics](https://docs.crewship.ai/observability/metrics.md): Scrape crewshipd's /metrics endpoint for process and domain metrics
- [Feedback](https://docs.crewship.ai/guides/feedback.md): Per-message thumbs / edit / regenerate signals bound to OTel trace_id — the continuous-learning loop for production agents.
- [Telemetry](https://docs.crewship.ai/guides/telemetry.md): Anonymous crash reporting — defaults by build flavour, the onboarding consent step, the Sentry DSN override, and exactly what does and does not leave your machine.
- [Chat interaction events](https://docs.crewship.ai/guides/chat-telemetry.md): The measurement behind the chat surface — the event vocabulary, what a payload may and may not contain, why these are not journal entries, and where the sink is wired.
- [Scheduling](https://docs.crewship.ai/guides/scheduling.md): Cron-based automated agent runs with custom prompts, timeout management, and execution tracking.
- [Files and Output](https://docs.crewship.ai/guides/files-and-output.md): Agent file system layout, file upload/download APIs, shared crew space, and real-time file events via WebSocket.
- [Paymaster](https://docs.crewship.ai/guides/paymaster.md): Cost ledger and hierarchical budget enforcement for every LLM call the platform makes.
- [Lookout](https://docs.crewship.ai/guides/lookout.md): In-process guardrails: prompt injection detection, tool-arg schema validation, structured output parsing, and secrets redaction.
- [Harbormaster](https://docs.crewship.ai/guides/harbormaster.md): Human-in-the-loop (HITL) approval queue that gates high-risk agent actions behind a human decision.
- [Cartographer](https://docs.crewship.ai/guides/cartographer.md): Mission checkpoints, forks, and advisory restore anchored to the journal cursor.
- [Lifecycle Hooks](https://docs.crewship.ai/guides/hooks.md): Shell, HTTP, or subagent callbacks that fire on platform lifecycle events. Blocking and non-blocking modes supported.
- [Automations](https://docs.crewship.ai/guides/automations.md): Run a routine whenever a journal event happens, without writing any glue.
- [Pages](https://docs.crewship.ai/guides/pages.md): Panel dashboards a routine or script pushes data into — permissioned per panel, never queried, and honest about when a number goes stale.
- [Operate Pages applications](https://docs.crewship.ai/guides/pages-apps-operations.md): Install, verify, back up and upgrade the offline Pages runtime.
- [Quartermaster](https://docs.crewship.ai/guides/quartermaster.md): Mission replay, regression detection, and LLM-as-judge evaluation over journal trajectories.
- [Watch Roster](https://docs.crewship.ai/guides/watch-roster.md): Live agent presence tracking: online, busy, blocked, offline. Sweeper marks agents offline after 5 minutes idle.
- [Backup & Restore](https://docs.crewship.ai/guides/backup.md): Admin-only workspace and crew backup bundles — AGE-encrypted .tar.zst with forward-compatible manifest, advisory locking, retention rotation, and dry-run restores.
- [Database Performance](https://docs.crewship.ai/guides/database-performance.md): How Crewship's SQLite layer behaves under concurrent agents — the write-lock model, WAL checkpointing, and what the measured ceilings actually are.
- [Docker Socket Proxy](https://docs.crewship.ai/guides/docker-socket-proxy.md): Run the server without raw /var/run/docker.sock: the exact Engine API endpoints Crewship uses, a compose deployment, and a straight account of what this control does not cover.
- [Service disk quotas](https://docs.crewship.ai/guides/service-disk-quotas.md): Run a crew service with a read-only root, bounded scratch space and fixed-size persistent volumes.
- [Migrations Catalog](https://docs.crewship.ai/guides/migrations.md): Index of every Crewship database migration from v50 onwards — what each one added, which PR introduced it, and whether it changes user-visible behaviour.
- [Migration Checklist](https://docs.crewship.ai/guides/migration-checklist.md): What to check before adding a schema migration — and which mistakes the build already catches for you.
- [Troubleshooting](https://docs.crewship.ai/guides/troubleshooting.md): First-line diagnostics for the most common Crewship issues — start with `crewship doctor`, then walk this page by symptom.
- [Upgrading & Downgrading](https://docs.crewship.ai/guides/upgrades.md): Keep crewship current with a single command, and roll back safely when an upgrade goes wrong.
- [Threat Model](https://docs.crewship.ai/security/threat-model.md): One-page summary of the Crewship trust boundaries: container isolation, sidecar credential broker, output scrubber, and the Keeper credential gate.
- [Roles & Permissions](https://docs.crewship.ai/security/rbac.md): 5-tier role-based access control (RBAC) for workspaces, agent roles, and the Crewship equality philosophy.
- [Credential Encryption](https://docs.crewship.ai/security/encryption.md): AES-256-GCM encryption with 16-byte IV, versioned keys, and Go/TypeScript cross-compatibility.
- [Container Isolation](https://docs.crewship.ai/security/container-isolation.md): How Crewship isolates agents using UID boundaries, capability dropping, network policies, and the sidecar proxy.
- [Audit Log](https://docs.crewship.ai/security/audit.md): Append-only, best-effort audit trail covering entity mutations and Keeper security decisions.
- [GDPR — Article 15 access + Article 17 cascade](https://docs.crewship.ai/security/gdpr.md): Operator playbook for honouring subject access requests (SAR) and right-to-be-forgotten (RTBF) using the admin GDPR endpoints + audit table.
- [Rate Limiting & Security Headers](https://docs.crewship.ai/security/security-headers.md): Per-IP HTTP rate limits on auth and API routes, standard security response headers, and single-use OAuth state tokens with 15-minute expiry.
- [Instance-Scope Backup Security](https://docs.crewship.ai/security/backup-instance-security.md): Threat model, crypto chain, recovery kit, quiet window and operational requirements for whole-instance backups.
- [Environment Variables](https://docs.crewship.ai/configuration/environment.md): Complete reference for all environment variables that configure the Crewship server, containers, auth, and integrations.
- [YAML Configuration](https://docs.crewship.ai/configuration/yaml-config.md): Configure Crewship via a YAML config file with server, container, storage, auth, Keeper, and LLM proxy settings.
- [Providers](https://docs.crewship.ai/configuration/providers.md): Container, storage, and state provider options including Docker, Apple Containers, and BoltDB.
- [Apple Containers on macOS](https://docs.crewship.ai/configuration/apple-containers.md): Running Crewship crews on Apple's container runtime: prerequisites, how provisioning builds instead of commits, build performance, and what this provider does not apply.
- [Devcontainer Configuration](https://docs.crewship.ai/configuration/devcontainers.md): Reference for per-crew devcontainer config fields, supported features, mise tools, and the sidecar bind-mount paths.
- [Manifest Schema](https://docs.crewship.ai/configuration/manifest-schema.md): Flat reference for every field in a Crewship workspace manifest (apiVersion crewship/v1).
- [Overview](https://docs.crewship.ai/manifest/overview.md)
- [Workspace](https://docs.crewship.ai/manifest/workspace.md)
- [Crew](https://docs.crewship.ai/manifest/crew.md)
- [Agent](https://docs.crewship.ai/manifest/agent.md)
- [Skill](https://docs.crewship.ai/manifest/skill.md)
- [Integration](https://docs.crewship.ai/manifest/integration.md)
- [Connector](https://docs.crewship.ai/manifest/connector.md)
- [Issue](https://docs.crewship.ai/manifest/issue.md)
- [Project](https://docs.crewship.ai/manifest/project.md)
- [Milestone](https://docs.crewship.ai/manifest/milestone.md)
- [Label](https://docs.crewship.ai/manifest/label.md)
- [Recurring issue](https://docs.crewship.ai/manifest/recurring_issue.md)
- [Triage rule](https://docs.crewship.ai/manifest/triage_rule.md)
- [Saved view](https://docs.crewship.ai/manifest/saved_view.md)
- [Workflow template](https://docs.crewship.ai/manifest/workflow_template.md)
- [Routine](https://docs.crewship.ai/manifest/routine.md)
- [Page](https://docs.crewship.ai/manifest/page.md)
- [Feature flag](https://docs.crewship.ai/manifest/feature_flag.md)
- [Instance setting](https://docs.crewship.ai/manifest/instance_setting.md)
- [Hook](https://docs.crewship.ai/manifest/hook.md)
- [Recipe](https://docs.crewship.ai/manifest/recipe.md)
- [Crew template](https://docs.crewship.ai/manifest/crew_template.md)
- [CLI Overview](https://docs.crewship.ai/cli/overview.md): Global flags, authentication, output formats, and top-level commands for the Crewship CLI.
- [Commands](https://docs.crewship.ai/cli/commands.md): Dump the full CLI command tree as a machine-readable manifest.
- [API](https://docs.crewship.ai/cli/api.md): Search API operations, inspect schemas, and call HTTP endpoints from the CLI.
- [AI clients](https://docs.crewship.ai/cli/ai.md): Connect AI clients to the MCP server and Agent Skill embedded in the Crewship binary.
- [crewship start](https://docs.crewship.ai/cli/start.md): Boot the Crewship server — Go binary, embedded Next.js UI, SQLite-by-default, optional Docker provider.
- [Onboarding](https://docs.crewship.ai/cli/onboarding.md): Propose a crew, inspect the proposal, and apply it under your own session.
- [Login, Logout, Whoami](https://docs.crewship.ai/cli/login.md): Authenticate the CLI against a Crewship server — interactive credentials, API token, Google OAuth, or device-code pairing — then check who you are.
- [crewship version](https://docs.crewship.ai/cli/version.md): Print the build identity of the local binary — and, with --remote, of the server it talks to.
- [crewship doctor](https://docs.crewship.ai/cli/doctor.md): Pre-flight diagnostic: container runtime, port binding, DB perms, telemetry status, update availability.
- [crewship paths](https://docs.crewship.ai/cli/paths.md): Inspect effective local installation paths, their sources and legacy layout conflicts without writing files.
- [crewship reset](https://docs.crewship.ai/cli/reset.md): Reset application data of a stopped local installation while retaining identity and management files.
- [crewship telemetry](https://docs.crewship.ai/cli/telemetry.md): Manage anonymous crash reporting consent — opt in, opt out, show current state and endpoint.
- [AI Workflows](https://docs.crewship.ai/cli/ai-workflows.md): Compose stdin, files, git context, and agents into shell pipelines.
- [Apply](https://docs.crewship.ai/cli/apply.md): Apply a workspace manifest. Idempotent, plan-then-confirm.
- [Export](https://docs.crewship.ai/cli/export.md): Export a crew, a page, or the whole workspace as a YAML manifest, or bundle a single run's chat and journal artefacts.
- [Init](https://docs.crewship.ai/cli/init.md): Bootstrap the first admin user on a fresh Crewship instance and print a CLI token.
- [Setup](https://docs.crewship.ai/cli/setup.md): Finish onboarding from the terminal — pick a crew template, an adapter, paste a CLI token.
- [auth](https://docs.crewship.ai/cli/auth.md): Self-service account management from the CLI — change your own password, pair another CLI.
- [Session](https://docs.crewship.ai/cli/session.md): Audit and revoke active browser sessions — the CLI mirror of Settings → Sessions.
- [Token](https://docs.crewship.ai/cli/token.md): Manage CLI authentication tokens — list, create, revoke, rotate, validate.
- [Me, Today, Now](https://docs.crewship.ai/cli/me.md): Three quick-action commands that compose multiple REST endpoints into single-screen workflow shortcuts — what's mine, what happened today, what's running now.
- [Admin](https://docs.crewship.ai/cli/admin.md): Host-side operator commands on the database file on this machine. The recovery surface for when a user (typically you) cannot log in — plus list-users, which reads the server the CLI targets.
- [logout](https://docs.crewship.ai/cli/logout.md): Remove stored CLI authentication credentials.
- [whoami](https://docs.crewship.ai/cli/whoami.md): Display the authenticated CLI identity and active workspace.
- [Agent Commands](https://docs.crewship.ai/cli/agent.md): Create, manage, and monitor AI agents with the Crewship CLI.
- [Ask](https://docs.crewship.ai/cli/ask.md): One-shot prompt to your default agent. Pipe-friendly, stdin-aware, supports fan-out across agents.
- [Chat](https://docs.crewship.ai/cli/chat.md): Inspect, attach to, and react inside chat sessions — the persistent thread behind every agent run.
- [crewship context](https://docs.crewship.ai/cli/context.md): Inspect authorized restricted context, immutable project files and private routine results.
- [conversation](https://docs.crewship.ai/cli/conversation.md): Search indexed conversation history from the CLI.
- [Feedback](https://docs.crewship.ai/cli/feedback.md): Record and inspect typed feedback signals (helpful, not_helpful, inaccurate, unsafe, edit, regenerate) on messages.
- [Run](https://docs.crewship.ai/cli/run.md): Run an agent with a prompt, stream the response, optionally hop into an interactive REPL.
- [Resume](https://docs.crewship.ai/cli/resume.md): Pick up an existing session by chat-id, run-id, or PR URL — or pick interactively from the 10 most recently active sessions.
- [Retry](https://docs.crewship.ai/cli/retry.md): Re-run a previous run with the same agent + recovered prompt — or with a tweaked prompt — into a new chat or continuing the old one.
- [Wait](https://docs.crewship.ai/cli/wait.md): Block until a run reaches a terminal status. kubectl-wait-style exit codes for shell scripting.
- [Watch](https://docs.crewship.ai/cli/watch.md): Top-level shortcut for `crewship journal --follow` — live SSE tail of the Crew Journal.
- [History](https://docs.crewship.ai/cli/history.md): Workspace-wide list of recent agent runs with timestamps, status, trigger, and (optionally) the prompt that kicked each run off.
- [Recall](https://docs.crewship.ai/cli/recall.md): Free-text FTS5 search across the Crew Journal — "did we discuss / try / decide X before?"
- [Recap](https://docs.crewship.ai/cli/recap.md): Agent-generated summary of a chat session — outcome, key decisions, open threads, and a suggested follow-up.
- [Inspect](https://docs.crewship.ai/cli/inspect.md): Deterministic structured timeline of a single run — journal events, cost roll-up, and tool-call counts (no LLM).
- [model](https://docs.crewship.ai/cli/model.md): Discover models available from a provider, and price a call before you make it.
- [crewship provider](https://docs.crewship.ai/cli/provider.md): Inspect the LLM providers this build can talk to, see how the agent sidecar routes each one, and send one live completion to prove a backend answers.
- [rehire](https://docs.crewship.ai/cli/rehire.md): Extend or resurrect an ephemeral agent.
- [crewship capacity](https://docs.crewship.ai/cli/capacity.md): What host admission control is holding, and why — the difference between a queued container start and a hung one.
- [Crew Commands](https://docs.crewship.ai/cli/crew.md): Create and manage crews, members, connections, and MCP configurations.
- [Project](https://docs.crewship.ai/cli/project.md): Group missions and issues under a delivery project. Status, priority, health, progress, breakdown stats.
- [Workspace](https://docs.crewship.ai/cli/workspace.md): Switch the active workspace, list members, invite users, and update workspace settings.
- [Mission Commands](https://docs.crewship.ai/cli/mission.md): Create, run, and manage missions and their tasks with the Crewship CLI.
- [Routine Commands](https://docs.crewship.ai/cli/routine.md): Manage workspace routines — declarative AI workflow recipes with cron + webhook triggers, two-tier execution, HITL waitpoints, and immutable version history.
- [Page Commands](https://docs.crewship.ai/cli/page.md): Create, permission, and feed data into Pages — panel dashboards a routine or script pushes to, never queries.
- [Pipeline Commands](https://docs.crewship.ai/cli/pipeline.md): Backwards-compatible alias for the routine CLI.
- [Routine Webhook Commands](https://docs.crewship.ai/cli/webhook.md): Manage event-driven webhook triggers attached to routines.
- [Routine Waitpoint Commands](https://docs.crewship.ai/cli/waitpoint.md): Inspect and decide human approval waitpoints created by routines.
- [Signal Commands](https://docs.crewship.ai/cli/signal.md): Wake every routine run parked on a named event, without knowing which runs they are.
- [Issue Commands](https://docs.crewship.ai/cli/issue.md): Issue tracking with labels, comments, and agent dispatch.
- [Label Commands](https://docs.crewship.ai/cli/label.md): Manage workspace-level labels — create, list, update, delete. Labels apply to issues across every crew.
- [Triage](https://docs.crewship.ai/cli/triage.md): Manage triage rules and run the triage processor over pending issues.
- [Recurring](https://docs.crewship.ai/cli/recurring.md): Manage recurring-issue schedules — scheduled issue creation.
- [Saved View](https://docs.crewship.ai/cli/saved-view.md): Manage saved views — query bookmarks across issues / runs / activity.
- [Escalation](https://docs.crewship.ai/cli/escalation.md): List, resolve, cancel, expire, and count crew escalations — when an agent flags something for human attention.
- [Workflow](https://docs.crewship.ai/cli/workflow.md): Manage workspace workflow templates — reusable issue-lifecycle stage definitions.
- [Skill Commands](https://docs.crewship.ai/cli/skill.md): Author, import, and assign skills to agents from the CLI.
- [Integration](https://docs.crewship.ai/cli/integration.md): Manage MCP server integrations at workspace, crew, and agent scope.
- [Recipe](https://docs.crewship.ai/cli/recipe.md): Browse and install crew recipes — 1-click crew + credentials + MCP templates.
- [Connector](https://docs.crewship.ai/cli/connector.md): Browse, verify, and install generic connectors from the manifest-driven catalogue.
- [MCP](https://docs.crewship.ai/cli/mcp.md): MCP registry, audit log, and raw mcp_config_json editing for crews and agents.
- [MCP Calls](https://docs.crewship.ai/cli/mcp-calls.md): Audit MCP tool invocations across the workspace — which tool, which agent, what arguments, what outcome.
- [Inbox](https://docs.crewship.ai/cli/inbox.md): Triage waitpoints, escalations, failed runs, and human-targeted messages from the terminal.
- [Notify](https://docs.crewship.ai/cli/notify.md): Desktop notifications — enable / disable / status / test / send. Crewship pings you when long runs finish.
- [Notification Channels](https://docs.crewship.ai/cli/notifychannel.md): Create, test, and manage outbound notification delivery channels.
- [Digest](https://docs.crewship.ai/cli/digest.md): Workspace digest routine — a token-zero ops summary of runs, cost, and failures, delivered via the notification system.
- [crewship memory](https://docs.crewship.ai/cli/memory.md): Inspect and search memory FTS5 indexes directly on the local filesystem.
- [Journal Commands](https://docs.crewship.ai/cli/journal.md): View the Crew Journal event stream from the command line.
- [Audit](https://docs.crewship.ai/cli/audit.md): Query the workspace audit log — by action, entity, user, date range, or free-text.
- [Privacy Commands](https://docs.crewship.ai/cli/privacy.md): GDPR self-service for peer memory — consent, your peer cards, and the operator model agents read about you.
- [Activity](https://docs.crewship.ai/cli/activity.md): Cross-crew activity feed — assignments, peer conversations, and escalations merged into one DESC-by-time list.
- [crewship chain](https://docs.crewship.ai/cli/chain.md): Follow what caused what across automations, issues, routines, runs and delegated assignments from a single anchor.
- [Logs](https://docs.crewship.ai/cli/logs.md): Tail an agent's container logs — recent lines or live WebSocket stream.
- [Metrics](https://docs.crewship.ai/cli/metrics.md): MISSION performance summary or a metric timeseries over a window. Not workspace-wide spend — that is crewship cost.
- [Presence Commands](https://docs.crewship.ai/cli/presence.md): Watch Roster — who is online, busy, blocked, or offline.
- [Work Commands](https://docs.crewship.ai/cli/work.md): Read the durable work ledger — what Crewship accepted, which attempt ran, and why it ended — then cancel or replay one item.
- [Cost](https://docs.crewship.ai/cli/cost.md): Quick spend summary — total, top spenders, per-crew rollup, subscription usage — in one screen.
- [Spend](https://docs.crewship.ai/cli/spend.md): Journal-native cost rollup — by agent, by routine, and top spenders.
- [Paymaster Commands](https://docs.crewship.ai/cli/paymaster.md): Cost and budget rollups across the workspace.
- [Approvals Commands](https://docs.crewship.ai/cli/approvals.md): List and decide Harbormaster human-in-the-loop approval requests.
- [Policy Commands](https://docs.crewship.ai/cli/policy.md): Manage per-crew autonomy and behavior-mode policy (PR-B F2).
- [Hire / Rehire Commands](https://docs.crewship.ai/cli/hire.md): Spawn a short-lived ephemeral agent into a crew (PR-D F5) and renew its TTL when the work isn't finished.
- [crewship persona](https://docs.crewship.ai/cli/persona.md): View, edit, reset, and inspect the version history of PERSONA.md for an agent or a crew default.
- [Hooks Commands](https://docs.crewship.ai/cli/hooks.md): Register, edit, remove, and toggle lifecycle hooks.
- [Automation Commands](https://docs.crewship.ai/cli/automation.md): Create and manage rules that run a routine when a journal event happens.
- [Checkpoint Commands](https://docs.crewship.ai/cli/checkpoint.md): Manage Cartographer mission checkpoints: create, list, restore (advisory), fork, delete.
- [keeper](https://docs.crewship.ai/cli/keeper.md): Configure, inspect, test, and audit Crewship's Keeper security controls from the CLI.
- [Config](https://docs.crewship.ai/cli/config.md): Manage the local CLI configuration file (~/.crewship/cli-config.yaml) — server URL, token, default workspace, and a one-shot health check.
- [Preferences Commands](https://docs.crewship.ai/cli/preferences.md): Read and write your personal, per-user UI preferences.
- [Server profiles](https://docs.crewship.ai/cli/server.md): Target several Crewship instances (dev/staging/prod) from one CLI config — named profiles, each with its own token, switchable per-command, per-shell, or per-directory.
- [System & Admin Commands](https://docs.crewship.ai/cli/system.md): crewship system — runtime & license info, Keeper status, admin stats, onboarding, and auxiliary-model assignment.
- [crewship instance](https://docs.crewship.ai/cli/instance.md): Read and write instance-wide settings stored in the singleton app_settings table.
- [Backup Commands](https://docs.crewship.ai/cli/backup.md): Admin-only backup and restore — create, inspect, verify, restore, rotate, and unlock encrypted workspace / crew bundles.
- [crewship recover](https://docs.crewship.ai/cli/recover.md): Restore a whole Crewship instance from an instance backup into an empty data directory — offline, with every automation held.
- [crewship features](https://docs.crewship.ai/cli/features.md): Browse the devcontainer features catalog the workspace can install on crew runtime images.
- [crewship feature-flag](https://docs.crewship.ai/cli/feature-flag.md): List feature flags, toggle per-workspace overrides, or drop overrides to inherit the instance default.
- [crewship runtimes](https://docs.crewship.ai/cli/runtimes.md): Browse the mise-managed runtime/tool catalog (Node, Python, Go, Terraform, ...) for pinning versions in a crew config.
- [Eval Commands](https://docs.crewship.ai/cli/eval.md): Two evaluation surfaces: mission replay/regression (Quartermaster) and routine-eval scenarios (cross-tier consistency suite, baseline regression diff).
- [Consolidate Commands](https://docs.crewship.ai/cli/consolidate.md): Force an immediate memory consolidation run.
- [self-update](https://docs.crewship.ai/cli/self-update.md): Upgrade the Crewship binary from a released installation.
- [Credential Commands](https://docs.crewship.ai/cli/credential.md): Manage encrypted credentials, assign them to agents, and validate API keys.
- [OAuth Commands](https://docs.crewship.ai/cli/oauth.md): Connect OAuth credentials headlessly and discover a provider's endpoints.
- [Seed](https://docs.crewship.ai/cli/seed.md): Provision a coherent small-business demo — crews, agents, credentials, routines, Pages, and executable issues — through the REST API.
- [Nuke](https://docs.crewship.ai/cli/nuke.md): Tear down workspace contents — DB entities, inbox items, escalations, and each crew's docker containers+volumes — without touching cached images.
- [db](https://docs.crewship.ai/cli/db.md): Local SQLite database maintenance commands.
- [Diff](https://docs.crewship.ai/cli/diff.md): Compare two persisted runs side-by-side — status, output, journal head.
- [now](https://docs.crewship.ai/cli/now.md): Show live workspace status from the CLI.
- [today](https://docs.crewship.ai/cli/today.md): Show today's runs and workspace spend.
- [Explain](https://docs.crewship.ai/cli/explain.md): Summarize what happened in a run — feeds journal entries to the default agent for a plain-English narrative.
- [Expose](https://docs.crewship.ai/cli/expose.md): Audit and revoke agent-initiated port exposures (capability URLs). The agent creates them via the sidecar; humans audit and tear down.
- [Lint](https://docs.crewship.ai/cli/lint.md): Static-analyse local CLI configuration and prompt library — catch the boring class of mistakes before they hit at use-time.
- [Open](https://docs.crewship.ai/cli/open.md): Open the web-UI page for a CLI resource in your default browser. Hybrid terminal/web workflow.
- [Plan](https://docs.crewship.ai/cli/plan.md): Ask an agent for a step-by-step plan; tool use is not technically restricted.
- [Prompt](https://docs.crewship.ai/cli/prompt.md): Local prompt library — save, list, view, edit, and delete prompts kept under ~/.crewship/prompts/. Pure local, no server round-trip.
- [Slash](https://docs.crewship.ai/cli/slash.md): User-defined slash commands loaded from ~/.crewship/commands/*.md and mounted as cobra subcommands at startup.
- [Shell](https://docs.crewship.ai/cli/shell.md): Interactive REPL — each line is a prompt to the default agent. Slash-commands control session state.
- [TUI](https://docs.crewship.ai/cli/tui.md): Real-time Bubble Tea dashboard — Mission Control for running missions, approvals, and the live journal.
- [Completion](https://docs.crewship.ai/cli/completion.md): Generate shell completion scripts for bash, zsh, fish, or powershell.
- [Copy Prompt](https://docs.crewship.ai/cli/copy-prompt.md): Recover the first user prompt from a previous run — to stdout or the system clipboard. Edit it, re-issue it.
- [Template](https://docs.crewship.ai/cli/template.md): Crew templates — blueprints for pre-configured crews. List, inspect, and deploy.
- [crewship decisions](https://docs.crewship.ai/cli/decisions.md): Opt-in Jev typed decisions for triage and retrieval experiments.
- [API Overview](https://docs.crewship.ai/api-reference/overview.md): REST API conventions, authentication, error handling, and communication layers for the Crewship platform.
- [OpenAPI Spec](https://docs.crewship.ai/api-reference/openapi.md): Machine-readable route catalog served at GET /openapi.json for API discovery and tooling.
- [Authentication](https://docs.crewship.ai/api-reference/auth.md): Email/password signup, Google sign-in, device pairing for the CLI, active-session management, long-lived CLI tokens, and email-based password recovery.
- [Onboarding](https://docs.crewship.ai/api-reference/onboarding.md): The conversational first-run: a setup agent proposes a crew, and only a human's apply creates it.
- [Workspaces](https://docs.crewship.ai/api-reference/workspaces.md): Manage workspaces -- the top-level tenant boundary that owns crews, members, invitations, skills, pipelines, schedules, and webhooks.
- [Crews](https://docs.crewship.ai/api-reference/crews.md): Manage crews -- isolated agent teams with their own containers, network policies, and members.
- [Agents](https://docs.crewship.ai/api-reference/agents.md): Manage agents — AI coding agents with configurable roles, CLI adapters, skills, and credentials. Agents in a crew share one Linux container.
- [Runs](https://docs.crewship.ai/api-reference/runs.md): Agent execution runs reconstructed from the Crew Journal stream.
- [Pipelines](https://docs.crewship.ai/api-reference/pipelines.md): Workspace-scoped pipeline and run API.
- [Routines](https://docs.crewship.ai/api-reference/routines.md): Scheduling and governance operations for saved pipelines.
- [Schedules](https://docs.crewship.ai/api-reference/schedules.md): Cron triggers for saved pipelines.
- [Waitpoints](https://docs.crewship.ai/api-reference/waitpoints.md): Human and external completion of paused routine steps.
- [Missions](https://docs.crewship.ai/api-reference/missions.md): Mission lifecycle management -- create, start, monitor, and control multi-agent task execution.
- [Issues](https://docs.crewship.ai/api-reference/issues.md): Issue tracker with projects, labels, comments, relations, and agent execution.
- [Projects](https://docs.crewship.ai/api-reference/projects.md): Manage workspace-scoped projects and their issue rollups.
- [Labels](https://docs.crewship.ai/api-reference/labels.md): Manage workspace-scoped issue labels.
- [Milestones](https://docs.crewship.ai/api-reference/milestones.md): Project milestones — named, dated checkpoints with issue rollup counts. List and create are nested under projects; update and delete take the milestone id directly.
- [Recurring Issues](https://docs.crewship.ai/api-reference/recurring-issues.md): Cron-scheduled issue templates that produce a fresh backlog row on every tick. CRUD plus server-side cron parsing and next_run recomputation.
- [Saved Views](https://docs.crewship.ai/api-reference/saved-views.md): Per-user saved issue views with optional workspace sharing — store filters_json + sort_json so a user (or the workspace) can replay a saved filter from one click.
- [Pages](https://docs.crewship.ai/api-reference/pages.md): Panel dashboards that producers push to and nobody queries — the page holds no datasource, no credential, and no clock of its own.
- [Triage Rules](https://docs.crewship.ai/api-reference/triage.md): Pattern-matching rules that auto-assign new backlog issues (crew, agent, project, priority, labels). CRUD plus a one-shot Process endpoint.
- [Workflow Templates](https://docs.crewship.ai/api-reference/workflow-templates.md): Reusable issue-lifecycle stage definitions (backlog → in_progress → done) attached to crews — workspace-scoped, with built-in and user-created rows sharing the same shape.
- [Credentials](https://docs.crewship.ai/api-reference/credentials.md): Encrypted credential vault for managing API keys, tokens, and secrets.
- [Skills & Templates](https://docs.crewship.ai/api-reference/skills.md): Skill marketplace, skill import, workflow templates, and crew templates.
- [Recipes](https://docs.crewship.ai/api-reference/recipes.md): 1-click crew bundles — list, preview (dry run), install. A recipe ships a curated crew + the credentials and MCP servers it needs; install commits everything in one transaction.
- [Integrations](https://docs.crewship.ai/api-reference/integrations.md): MCP server integrations at workspace, crew, and agent levels, plus OAuth flow.
- [Connectors](https://docs.crewship.ai/api-reference/connectors.md): Curated connector catalog -- browse, verify, and install pre-built MCP integrations.
- [Conversations](https://docs.crewship.ai/api-reference/conversations.md): Message reactions, group-chat participants, mid-turn steering, and chat attachments — the read/write surface around chat sessions beyond the basic message list.
- [User Preferences](https://docs.crewship.ai/api-reference/preferences.md): Per-user JSON-blob key-value store backing UI settings (panel sizes, last-opened tabs, density, …). Each key is opaque to the server; the frontend owns the schema.
- [Privacy](https://docs.crewship.ai/api-reference/privacy.md): GDPR primitives for per-user peer memory -- list, purge, and opt out of peer cards, and read or correct the operator model.
- [Outbound notifications](https://docs.crewship.ai/api-reference/notification-channels.md): API reference for notification channels, agent pairing, and workspace notification templates.
- [Inbox](https://docs.crewship.ai/api-reference/inbox.md): Unified human-in-the-loop inbox — list, count, and transition items between unread / read / resolved. Backed by inbox_items (migration v85).
- [Webhooks](https://docs.crewship.ai/api-reference/webhooks.md): Inbound webhook endpoint for triggering agent execution from external systems.
- [Work Items](https://docs.crewship.ai/api-reference/work-items.md): Read the durable work ledger — what was accepted, which attempt is running, and why it ended — then cancel or replay one item.
- [WebSocket](https://docs.crewship.ai/api-reference/websocket.md): Real-time channels for log streaming, chat, events, and live updates.
- [Admin](https://docs.crewship.ai/api-reference/admin.md): Workspace administration endpoints for stats, user management, Keeper audit logs, and the license system.
- [Keeper administration](https://docs.crewship.ai/api-reference/keeper.md): Configure, test, and operate the Keeper credential-governance service.
- [Feature Flags](https://docs.crewship.ai/api-reference/feature-flags.md): Instance-global feature toggles with percentage rollout and per-workspace overrides.
- [Instance Settings](https://docs.crewship.ai/api-reference/instance-settings.md): CRUD over the instance-global key/value settings store, with redaction of sensitive values and protection of bootstrap markers.
- [Policies](https://docs.crewship.ai/api-reference/policies.md): Per-crew autonomy posture and behavior mode that gate every HITL-relevant decision the orchestrator makes.
- [Backup](https://docs.crewship.ai/api-reference/backup.md): REST endpoints for creating, listing, inspecting, verifying, restoring, rotating, and deleting backup bundles. Each route acts on the workspace named and admits its OWNER or ADMIN, or an instance administrator.
- [Provisioning](https://docs.crewship.ai/api-reference/provisioning.md): Devcontainer features and runtimes catalog, crew provision lifecycle, and the cache image registry.
- [Runtime Capacity](https://docs.crewship.ai/api-reference/runtime.md): Inspect host admission control and container starts held for capacity.
- [Port Expose](https://docs.crewship.ai/api-reference/port-expose.md): Capability URLs for exposing in-container ports to the operator's browser without modifying the agent's environment.
- [Journal](https://docs.crewship.ai/api-reference/journal.md): Read and stream the Crew Journal -- the canonical append-only event log.
- [Chains](https://docs.crewship.ai/api-reference/chains.md): Reconstruct the causal graph around one issue, run, routine, assignment or automation in a single call.
- [Paymaster](https://docs.crewship.ai/api-reference/paymaster.md): Cost rollups across crew, agent, and mission scopes.
- [Approvals](https://docs.crewship.ai/api-reference/approvals.md): Harbormaster human-in-the-loop approval queue.
- [Checkpoints](https://docs.crewship.ai/api-reference/checkpoints.md): Cartographer mission checkpoints: list, create, restore (advisory), fork, delete.
- [Hooks](https://docs.crewship.ai/api-reference/hooks.md): Lifecycle hook registry -- list, register, update, delete, and toggle.
- [Automations](https://docs.crewship.ai/api-reference/automations.md): Rules that enqueue a routine run when a matching journal event commits.
- [Feedback](https://docs.crewship.ai/api-reference/feedback.md): Per-message feedback signals (thumb-up/down, edit, regenerate) bound to OTel trace_id.
- [Presence](https://docs.crewship.ai/api-reference/presence.md): Watch Roster -- live agent status across a workspace or crew.
- [Memory](https://docs.crewship.ai/api-reference/memory.md): Memory health snapshot — five metrics that quantify how usable the episodic memory is.
- [Eval](https://docs.crewship.ai/api-reference/eval.md): Mission replay and regression detection via Quartermaster.
- [Consolidate](https://docs.crewship.ai/api-reference/consolidate.md): Trigger manual memory consolidation runs.
- [Slash Commands](https://docs.crewship.ai/api-reference/slash-commands.md): Server-driven slash command registry, capability-filtered to the caller's grants.
- [Internal IPC](https://docs.crewship.ai/api-reference/internal.md): Sidecar / orchestrator endpoints under /api/v1/internal/. X-Internal-Token authenticated, never exposed to agents.
- [Changelog](https://docs.crewship.ai/changelog/overview.md): What's new in Crewship — selected release highlights, with the upgrade notes that go with them.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.