Many Agents. One place to manage them.
BootAgent is a local desktop workspace for AI coding Agents. It helps you detect what is already on your machine, connect a Provider, configure an Agent, and get to a first successful request without editing several tool-specific files by hand.
| You want to... | Start with... |
|---|---|
| Try an Agent for the first time | Download a release, then follow the AI Agent Kit |
| Keep existing Agents in one place | Open Environment overview and let BootAgent scan the machine |
| Find a Skill, MCP server, plugin, or AI product | Open Marketplace, search or filter, then open an item for its details and upstream links |
| Move a setup to another machine | Open Settings > Import and export and choose a v1 JSON file or a v2 bundle |
| Fix an install or update | Open Task Center for the source, progress, and diagnostic output |
The shortest first-run path is:
Launch BootAgent
→ Choose or add a Provider
→ Create an API Key in that Provider's official account page
→ Choose an Agent and a Profile
→ Run setup
→ Send a first request
- Detect, install, update, launch, and uninstall supported CLI and desktop Agents.
- Detect Agents installed outside BootAgent when their command or known installation path is available.
- See long-running install and update tasks in Task Center, including progress, source, and cancellable steps.
- Uninstall selected installation instances while keeping Profiles, Providers, configuration files, and conversations.
- Use built-in Providers or add an OpenAI-compatible or Anthropic-compatible endpoint.
- Check an endpoint using the protocol that the selected Agent actually uses.
- Save reusable Profiles and select the Profile and model from an Agent's configuration screen.
- Keep API Keys in private local storage; they are not placed in ordinary summaries or recommendation prompts.
- Keep a local Skill library, scan Skills already present on supported Agents, and choose which Agents receive each Skill.
- See which models on your keyed Providers can understand images, video or audio, or generate images, and where each finding comes from (declared by the Provider, inferred from the model name, or probed). Listing models is free; a probe is a real request and is sent only when you click it, and an image-generation probe asks first.
- Generated images, audio and video default to
~/.bootagent/output/, outside the installed Skill directory, and never overwrite existing files. Use--outto choose another destination. - When BootAgent cannot confirm how a model is called, as with many user-added gateways, you can install the Skill in explore mode. The Agent then asks you or reads the provider's documentation to work out the request, asks before anything billed, and records what worked under
~/.bootagent/skill-notes/. Its script sends only to that provider's own host and adds the key itself. - Install a multimodal Skill (image understanding, image generation, speech-to-text, text-to-speech, video generation, video understanding) into Claude Code, Codex or OpenCode, with example prompts to copy. Agents without a Skills directory get the same files under
~/.bootagent/multimodal/and a prompt that points them there. Video understanding samples frames with ffmpeg, which must be onPATH. Built-in Providers add the models their docs name: MiniMax Speech on Novita, PPIO and JieKou.AI, GPT Image and async text-to-video on JieKou.AI. Video is never probed, because a probe would be a full billed generation. Its script needs Node.js 18 or later. BootAgent writes the Provider key into the Agent's own configuration asBOOTAGENT_<PROVIDER>_API_KEY, or into a private file under~/.bootagent/skill-env/for Agents with no configurable environment, so the Agent and its model can read it; the install confirmation says so. Uninstalling the last Skill that uses a Provider removes the variable. - Discover MCP servers from initialized Claude Code, Codex, OpenCode, Kilo CLI, and Hermes installations.
- Store MCP servers in a local user-level Registry, select sync targets explicitly, and apply changes only when you are ready.
- Bootstrap Node.js, uv, and Aider's managed Python runtime when an Agent needs them.
- Optionally run local API format conversion and launch BootAgent at login; both features are off by default.
- Migrate existing Codex and ChatGPT Desktop conversations into BootAgent's local provider bucket. This migration intentionally creates no history backup.
- Browse Skills, MCP servers, plugins, standalone AI products, prompt collections, and workflow templates.
- Combine category, source, use-case, API-key, and multi-type filters. An item can belong to more than one type.
- Refresh supported public sources independently. A source failure keeps the bundled catalog available instead of blanking the market.
- Open a detail page for the description, install guidance, README when the source provides one, and upstream links.
- Ask an installed Codex or Claude Code CLI to shortlist tools for a stated need. Returned IDs are checked against the current catalog before display.
- Save recommendation requests and result snapshots locally so a previous search can be reopened later. Recommendation history is never uploaded as telemetry.
- Import and export Providers, Profiles, selected MCP servers, and Skills.
- Legacy v1 JSON files remain importable. The v2 portable bundle separates configuration JSON from a Skills ZIP.
- Skills first enter BootAgent's managed library; you choose target Agents from the Skills page afterward.
- Preview new, overwritten, and conflicting resources before applying an import. Writes use snapshots and rollback on failure.
- API Keys and MCP secrets are omitted by default. Password-encrypted export or explicitly confirmed plaintext export is available.
BootAgent supports detection, configuration, launch, or installation guidance according to each Agent's upstream contract. An entry below does not mean that BootAgent redistributes the Agent package; the detail page shows the available install source.
| CLI and local Agents | Desktop Agents |
|---|---|
| Codex · Claude Code · OpenCode | DeepSeek Harness · Claude Desktop |
| Kilo CLI · Aider · OpenClaw | ChatGPT Desktop · WorkBuddy |
| Hermes Agent · Kimi Code · Pi | WorkBuddy AI · ZCode |
| DeepSeek Harness (local web app) | Kimi Code Desktop |
Claude Desktop can be detected and launched on macOS and Windows, but it must be installed by the user from the official download page. BootAgent writes compatible configuration and does not download or proxy Claude Desktop.
Kimi Code Desktop is installed automatically on macOS and Windows. Before running the installer, BootAgent checks the download's digest and the app's signature: the Developer ID team on macOS, and the Authenticode publisher on Windows. It shares ~/.kimi-code/config.toml with the Kimi Code CLI, so both Agents read the same configuration.
DeepSeek Harness can use its shipped DeepSeek route or a configured compatible Provider. It opens a local web app after installation; its own onboarding commands remain the source of truth.
JieKou.AI, PPIO, Novita, DeepSeek, and Moonshot are built-in Providers. Custom OpenAI-compatible and Anthropic-compatible Providers can be added from the Provider page.
Download the latest package from GitHub Releases.
| Platform | Recommended package | Architectures |
|---|---|---|
| macOS | DMG | Intel and Apple Silicon |
| Windows | NSIS installer | x64 and ARM64 |
| Linux | AppImage, deb, or rpm | amd64 and arm64 |
Linux also provides an OTA ZIP archive. Release artifacts include SHA256SUMS. macOS packages are Developer ID signed and notarized. The release channel is still a Wails Alpha technical preview: Windows and Linux packages are currently unsigned and may require an operating-system approval on first launch.
BootAgent does not redistribute Agent packages and does not bundle Node.js, Git, WebView, or API Keys. Missing prerequisites are reported with a link to the relevant official installation instructions.
- BootAgent is local-first. Provider credentials, configuration, recommendation history, and backups stay on the machine by default.
- Recommendation prompts contain the stated need and public catalog metadata only; the recommendation process receives no install or file-write tools.
- The latest three historical versions are kept for each Profile, Provider, MCP, Agent configuration target, and Skill by default. Backups live under
~/.bootagent/backup; retention can be changed in Settings. - Uninstall removes the selected program instance, not the user's Profiles, Providers, configuration files, or conversations.
- BootAgent is not a VPN, proxy, shared-key service, or Agent package distributor. Downloads use official sources, authorized mirrors, or documented manual installation paths.
| Symptom | First check |
|---|---|
| Agent is not listed | Refresh Environment overview, check that the command is on PATH, or use the Agent's documented manual install path. |
| Provider connection check fails | Confirm the endpoint and model match the Agent's protocol. Use Check model for a saved Profile; saving or applying a Profile does not automatically send a model request. |
Codex reports does not support endpoint: responses |
Codex is configured to use Responses. Select a model that supports it, or use the API protocol adapter with a working Chat Completions Profile and apply its generated Codex Profile. A model appearing in /models does not prove Responses support. |
| Windows detects a desktop app but its version is unavailable | BootAgent retries the registered package's metadata through Windows PackageManager. If the metadata remains inaccessible, launch is still available; this warning is separate from model API compatibility. |
| Marketplace looks incomplete | Check the source status and refresh. A bundled snapshot remains available when an online source is unavailable. |
| Import reports a conflict | Review the preview, choose skip or overwrite, and apply only after checking the affected resource names. |
| BootAgent cannot check for an update | In Settings, switch between GitHub and the domestic Gitee mirror, then retry. Existing installation files are not replaced by a failed check. |
| The Agent will not download | Follow the official install page or install manually, then return to BootAgent to detect the local installation. BootAgent does not bypass network restrictions. |
For users, the release packages above are the supported installation path. For development, you need Go (the version declared in go.mod), Node.js, pnpm 11.21.0, and the target platform's WebView dependencies. Linux builds use GTK4 and WebKitGTK 6.0.
git clone https://github.com/MaimoryLab/BootAgent.git
cd BootAgent
cd frontend
pnpm install --frozen-lockfile
pnpm run build
cd ..
go run -tags wails ./cmd/bootagent-desktop
For a production binary on the current host, use task build:desktop. Useful checks are:
go test ./...
go test -race ./...
go vet ./...
cd frontend && pnpm run test && pnpm run build
- AI Agent Kit: guided setup for a first request
- Documentation: specifications and architecture decisions
- Security policy
- Contributing
- Public site repository
- Issues and feature requests
BootAgent is released under the Apache License 2.0. Third-party notices are listed in NOTICE.