# AgentLayer - [What it is](/docs): AgentLayer turns Claude Code or Codex into an agent that remembers your work, in a folder you own. - Start here - [Quickstart](/docs/getting-started/quickstart): One line starts a setup guide that checks your machine and walks you through each step, then setup asks a few questions. - [Your first ten minutes](/docs/getting-started/first-ten-minutes): A first session that needs no keys. Say hello, teach the agent something, and see the next session start with it. - [How it works](/docs/getting-started/how-it-works): The home, the brain, the rails, and skills and packs. What each one is and where to read more. - [What setup asks](/docs/getting-started/onboarding): The questions setup asks, the files it writes, and the one rule every session depends on. - [Install by hand](/docs/getting-started/install): Every step the Quickstart runs, one at a time. For Windows, for starting with Codex, and for seeing each step. - Use it - [Daily rhythm](/docs/use/daily-rhythm): What to run and when. Briefings and the pulse, goals from the day to the year, attention for the day, standup, session checkpoints, and sync. - [Plan and run](/docs/use/plan-and-run): The work. Projects, tasks, the roadmap, the goals ladder, briefings, and the session radar. - [Build and ship](/docs/use/build-and-ship): The code. Engineering playbooks, worktrees, pull-request review, API collections, deploy checks, plan specs, and releases. - [Reach and see](/docs/use/reach-and-see): The world. The browser, audio and video, the database, web search, Slack, the support desk, the SEO suite, and, with a deployment, logs and uptime. - [Brain and memory](/docs/use/brain-and-memory): The context. Capture, compile, self-review, sync, the dashboard, seed bundles, and onboarding. - [Attention](/docs/use/attention): Plan your day and keep it in view. Today's three, what slipped, the week and month, the roadmap in flight, who is waiting on you, your standup, and a checklist of where every session stands. - [Proposals](/docs/use/proposals): Sessions the agent lines up for you to start. One long-running session, the heartbeat, keeps the best five waiting, and you choose Start, Later or Skip. - [Standup and checkpoints](/docs/use/standup-and-checkpoints): Where things stand, at a glance. Your standup as a card to read while you talk, and a checklist of where one session, or every recent session, has got to. - [Tasks](/docs/use/tasks): Markdown-native task management. One file per task, YAML frontmatter, threads as callouts, a generated board. - [Capture](/docs/use/capture): Feed the brain anything. Thoughts, files, URLs, clipboard, and corrections, from chat or the terminal. - [Sync](/docs/use/sync): One pass through every maintenance step, in dependency order, ending in a status block and, when something needs you, a few closing questions. - [Self-evolution](/docs/use/self-evolution): Corrections become learnings that load every session, and a monthly review, run with you, prunes the brain and turns recurring signals into prompt and skill edits. - [Tax and books](/docs/use/tax): Know what tax you owe at every moment, for your companies and your personal return. Every obligation becomes a dated task, receipts land in one ledger, the month closes for your accountant, and the estimate is planned before the penalty, not after. - Dashboard - [Dashboard](/docs/dashboard): A self-contained mission-control page, regenerated from your home, that shows your day, your work, and what the agent knows and can do on one screen. - [Today](/docs/dashboard/today): Your day, opened on Attention. What today is for, what slipped, the week and month, the roadmap in flight, who is waiting on you, what moved today, your goals, and the morning's news. - [Tasks and projects](/docs/dashboard/tasks-and-projects): Every open task by deadline, what's in flight, blocked, or going stale, and each project's progress, read straight from task frontmatter. - [Sessions](/docs/dashboard/sessions): What you and the agent worked on, captured automatically. The latest radar, 30 days of history, and where the work happened. - [Brain](/docs/dashboard/brain): What the agent loads every session, what it remembers, the threads it tracks, its concept articles, and the health of the compile pipeline. - [Reports and scheduler](/docs/dashboard/reports-and-scheduler): Everything the agent produced, by day, and the recurring routines that keep the home current. - [Capabilities](/docs/dashboard/capabilities): Everything you can ask the agent to do. A cheatsheet of what to say, every skill and tool, the CLI, the mods, and the hooks. - [Persona and system](/docs/dashboard/persona-and-system): Who the agent is and who it works for, its settings and logs, the changelog against your baseline, and the health checks behind the badge. - Workstation - [The rig](/docs/workstation): Optional. Running agents well is less about the model and more about the terminal, the keybindings, the worktrees, and where the brain lives. - [Terminal](/docs/workstation/terminal): Ghostty, the keybindings that pay for themselves, and the one Claude Code flag to turn on. - [cmux](/docs/workstation/cmux): A native-Mac terminal built on Ghostty for running many agents at once, with the AgentLayer sidebar showing who is waiting and what each agent is working on. - [Editor and tools](/docs/workstation/editor-and-tools): VS Code as the viewer, Obsidian as the brain reader, MarkEdit for documents, Tower for git review. - Engineering - [Engineering](/docs/engineering): How your agent works on code. A playbook per kind of task, 23 named principles, and review, verification, and release loops that never touch your remote on their own. - [The engineer skill](/docs/engineering/engineer): One skill routes every kind of code task to a playbook, pull requests included, applies the engineering principles, and runs a comment pass before you see the diff. - [Principles](/docs/engineering/principles): Twenty-three engineering principles, each a short rule with a test the agent runs on its own work. Name one to steer. - [Design and review](/docs/engineering/design-and-review): How much design a change deserves, and the tools for each rung: architect, arena, swarm, interrogate, and a second model on your work. - [Second-model review](/docs/engineering/second-model-review): Put a different model on anything you made, a proposal, a plan, a document, or code. It documents what it finds on one dossier; your agent verifies every finding, fixes the real ones, and hands back the next round. - [Pull requests](/docs/engineering/pull-requests): Review a teammate's PR, answer the review on yours, present yours, and sweep everything on GitHub waiting on you. You paste every comment; only your own PR's description is applied for you, once you approve it. - [Worktrees](/docs/engineering/worktrees): Parallel agents on one repo with zero collisions, and a skill that creates and bootstraps the checkout for you. - [Specs and plans](/docs/engineering/specs-and-plans): Turn an intent into a spec before code, design a bespoke workflow for large work, and keep a decision trail you can audit later. - [Coding rules](/docs/engineering/coding-rules): Where your agent's coding standards live, how language rules apply by path, how a repo's own manual takes precedence, and the default test policy. - [Verification](/docs/engineering/verification): What counts as proof. The proof ladder, a generated verification skill per repo, forensics, blast radius, and hillclimbing a metric. - [API collections](/docs/engineering/api-collections): Your agent drafts API requests as collection files you open and fire in your own client. It authors; you send. - [Releases](/docs/engineering/releases): How a change becomes a version. The release act, forward-only history, and what your agent will and won't do with your git remote. - Connect it - [Packs](/docs/packs): A pack connects your agent to one outside system, such as GitHub, a database or Slack. Each one ships with the plugin and stays off until you switch it on. - [GitHub](/docs/packs/github): Pull requests, issues and CI runs through eighteen tools, on one token per repo owner. Read-only unless your token allows a write, and nothing merges or closes. - [Database](/docs/packs/database): Read-only Postgres queries and schema for the databases you name, plus a scratch copy of a local database for risky schema work. - [Browser](/docs/packs/browser): Screenshots, styled PDFs, pages as Markdown, recordings and scripted flows on a browser the plugin installs for you, plus web search with an optional key. - [SEO](/docs/packs/seo): Search Console, PageSpeed, SerpAPI and Open Page Rank behind one read-only seo skill, plus wordpress-rest for site content. - [Slack](/docs/packs/slack): Read-only Slack through your own account. A pulse across your channels, the replies you owe, threads where someone outside the workspace is waiting, and channels you joined that the agent doesn't know about yet. It never posts and never reads DMs. - [Helpdesk](/docs/packs/helpdesk): Your support queue in SupportCandy or Pylon, worked the way your team works it. The agent reads, sorts and drafts; a person sends every reply. A queue profile per project holds its own pulse, bug lifecycle, routines and context. - [App Store](/docs/packs/app-store): Any app's App Store reviews, newest first, from Apple's public feed. No key and read-only; the agent never replies to a review. In a helpdesk queue, the bugs they report become cards beside its tickets. - [Xcode](/docs/packs/xcode): Build, test, run and debug Apple apps and Swift packages from a session, through the MCP server that ships with Xcode. macOS only, no key. - [API](/docs/packs/api): API requests drafted as collection files you open and send from your own client. No key, no tools and no permissions to grant. - [Accounts and costs](/docs/packs/accounts): The core needs zero external accounts. Packs add optional services, most of them free. - Run it - [Updating](/docs/run/updating): Pull a new plugin version, then reconcile your home with the upgrade skill. - [Hosts](/docs/run/hosts): Your agent runs inside an agent CLI you already use. The home is the same on every host; what differs is how the host is wired to it. - [Claude Code](/docs/run/claude-code): The first host. The version the plugin needs, what it adds to a session, and the settings a home sets for you. - [Codex](/docs/run/codex): Run the same home from OpenAI Codex. What to install, how to wire a home for it, what to trust, and the limits that remain. - [Configuration](/docs/run/configuration): The settings a home reads, the file each one goes in, and where every key lives. - [Adding another agent](/docs/run/another-agent): A second agent takes about five minutes of setup, because your computer is already set up. Step by step, no coding needed. - [Billing per agent](/docs/run/billing): Put each agent on its own Claude account, so personal and work usage never share a plan. Step by step, no coding needed. - [Seed bundles](/docs/run/seed-bundles): Hand an agent's persona and setup to a teammate as a zip. Memory, sessions, and credentials stay the recipient's own. - [History](/docs/run/history): Local version history for the agent home, turned on in one question, so any change to what the agent knows can be seen and undone. - [Naming your agent](/docs/run/naming): Your agent wakes up as Oppi. Call it whatever you like; the plugin stays agentlayer. - [Voice input](/docs/run/voice): Speak to your agent with the dictation already built into your computer. No install, no coding needed. - [Bring your Claude history](/docs/run/claude-history): Move what the Claude app already knows about you into your agent, so it doesn't start from zero. Step by step, no coding needed. - [Privacy](/docs/run/privacy): All data stays local. Secrets are deny-gated. Transcripts are redacted before they persist. - Reference - [How the brain works](/docs/reference/the-brain): Raw conversations are source code, the model is the compiler, and the compiled wiki is your queryable second brain. - [Agent home](/docs/reference/agent-home): One folder holds everything that makes the agent yours. What lives where, what loads at session start, and how the home is found. - [Architecture](/docs/reference/architecture): The brain is portable markdown. The host is the runtime. The plugin is the glue. - [Skills](/docs/reference/skills): All 36 skills, grouped the way the work is. Related work is one skill with playbooks, and you ask in plain words. - [MCP tools](/docs/reference/mcp-tools): The 84 tools the plugin's MCP server exposes, group by group, with what each does and whether it is granted, pack-gated or always asks. - [CLI](/docs/reference/cli): The plugin's CLI runs the MCP server's logic from a plain terminal, for cron jobs, scripts, and one-shots. - [Hooks](/docs/reference/hooks): The four moments the plugin hooks into on either host, what the shell guard refuses, and what capture redacts. - [Mods](/docs/reference/mods): The plugin's in-process module for Claude Code. Instant commands, queued prompts, this conversation's reports, repo manuals on first touch, sync progress, the heartbeat's clock, and notices above the prompt. - [Notices](/docs/reference/notices): The one list of upkeep only you can start, how loud each item gets, and where every surface shows it. - [Upgrades and releases](/docs/reference/upgrades): How the upgrade skill reconciles a home after a plugin update, what each version asks of you, and the format maintainers write. - [Changelog](/docs/changelog) - About - [FAQ](/docs/about/faq): Short answers to the questions that come up first. - [Platforms and limits](/docs/about/platforms): Built and tested on macOS. Linux and Windows work with a few caveats. Some limits are by design. - [Before the plugin](/docs/about/history): The agent started as a daemon with a chat bot and scheduled jobs. Subscription billing rules turned it into a plugin. - [Contributing](/docs/about/contributing): Pull requests welcome. Open an issue first for anything architectural.