# actual — AI context for your codebase

> actual is a CLI tool that analyzes a git repository, fetches the team's Architectural Decision Records (ADRs) from the Actual AI bank, tailors them to the detected languages and frameworks, and writes AI context files — keeping every coding agent architecturally aligned, automatically.

Install via Homebrew: `brew install actual-software/actual/actual`
Zero-install: `npx @actualai/actual adr-bot`

Output formats: CLAUDE.md (claude-md), AGENTS.md (agents-md), .cursor/rules/actual-policies.mdc (cursor-rules)

Supported AI backends (runners): claude-cli (default), anthropic-api, openai-api, codex-cli, cursor-cli

## Commands

- `actual adr-bot` — Analyze repo, fetch ADRs, tailor, and write the output file (main command)
- `actual auth` — Check Claude Code authentication status
- `actual status` — Show current config and state of managed output files in the current repo
- `actual config` — View or edit the global config file (~/.actualai/actual/config.yaml)
- `actual runners` — List all available AI backends and their current availability status
- `actual models` — List known model names grouped by runner
- `actual cache` — Clear local analysis and tailoring cache

## Key Flags (actual adr-bot)

- `--dry-run` — Preview changes without writing files
- `--force` — Skip confirmation prompts and bypass all caches
- `--output-format <FMT>` — claude-md (default) | agents-md | cursor-rules
- `--runner <RUNNER>` — AI backend: claude-cli | anthropic-api | openai-api | codex-cli | cursor-cli
- `--model <MODEL>` — Override AI model; runner is auto-inferred from the model name
- `--project <PATH>` — Target a sub-project in a monorepo (repeatable)
- `--no-tailor` — Write raw ADRs without AI tailoring

## Runners

- `claude-cli` (default) — requires the claude binary; install: `npm install -g @anthropic-ai/claude-code && claude auth login`
- `anthropic-api` — requires ANTHROPIC_API_KEY; set: `export ANTHROPIC_API_KEY=sk-ant-...`
- `openai-api` — requires OPENAI_API_KEY; set: `export OPENAI_API_KEY=sk-...`
- `codex-cli` — requires the codex binary; install: `npm install -g @openai/codex && codex login`
- `cursor-cli` — requires the agent binary; install: `curl https://cursor.com/install -fsS | bash`

## Output Formats

- `claude-md` → CLAUDE.md (Claude Code)
- `agents-md` → AGENTS.md (Codex CLI, OpenCode)
- `cursor-rules` → .cursor/rules/actual-policies.mdc (Cursor IDE)

## AI Skill

The actual AI skill teaches coding agents how to run, configure, and troubleshoot the CLI autonomously.

Install methods:
- `npx skills add actual-software/actual-skill` — one-command install for Claude Code
- `claude mcp add-plugin https://github.com/actual-software/actual-skill` — Claude Code plugin
- `clawhub install actual` — ClawdHub / OpenClaw

Source: [actual-software/actual-skill](https://github.com/actual-software/actual-skill) (Apache-2.0) | [actual-software/actual-skill-openclaw](https://github.com/actual-software/actual-skill-openclaw) (MIT-0)

## Docs

- [Getting Started & Full Command Reference](https://cli.actual.ai/docs.md): Installation, quick start, all commands with flags, runners, output formats, configuration, and troubleshooting

## Optional

- [Product overview](https://cli.actual.ai/index.md): Homepage — tagline, install instructions, how it works
- [actual.ai](https://actual.ai): Main product site with pricing, about, blog, and agent documentation