ClickClackClaw
github.com/openclaw/clickclack- Category
- Social
- Rank
- No. 1347Tools index
Previous survey · No. 1359 ·
- Pricing
- Open Source
- Type
- APP
- Builder
- openclaw
- GitHub
- 201 stars
- Latest release
- v0.4.1
- Date
About
The chat app with claws — OpenClaw-native messaging built around agent participation as first-class users.
What it does
ClickClack combines team messaging, administration, automation, and bot integration in one service. People use channels, threads, direct messages, search, uploads, and moderation through web or desktop clients. Scripts and bots connect through the command line, TypeScript client, REST API, or recoverable WebSocket event stream.
Why it's ranked here
This is a credible choice for teams that want self-hosted chat with serious automation surfaces. One server binary embeds the web application and migrations, while cursor-based event recovery gives integrations a durable foundation. Its broad client and storage options are strong, but the unfinished V1 scope and missing dedicated SDK tests temper that confidence.
What's good
The operational shape stays compact: SQLite and local uploads work by default, while Postgres and R2 provide expansion paths. Realtime clients can recover missed events using cursors. Bot tokens are scoped, raw tokens appear only when created, and historical authorship survives bot deletion. The command line also handles migrations, hot SQLite backups, exports, administration, and scripted chat.
Tradeoffs
V1 remains in progress. Multi-node WebSocket fanout, federation, voice, video, and full Mattermost compatibility are outside the stated scope. The TypeScript client has no dedicated test target, relying on a bot example as its smoke test. Production self-hosting also requires explicit authentication, storage, TLS termination, and backup configuration.
How to use it well
Use ClickClack for small teams, internal tools, or communities that need human conversation and programmable bot participation in the same workspace. Start with the embedded SQLite deployment, automate administration and chat through the command line, then move storage outward if needed. Choose another adjacent system if voice, video, federation, or multi-node realtime fanout is essential.
Technical notes+
README.md defines the product surfaces and single-binary deployment model. go.mod identifies the Go module and includes Chi, WebSocket, SQLite, Postgres, JWT, OAuth, and YAML dependencies. package.json coordinates web, SDK, desktop, lint, type checking, Go tests, Playwright, formatting, embedding, and Cloudflare deployment tasks through pnpm. packages/sdk-ts/src/index.ts contains framework-neutral HTTP and realtime client types, bot lifecycle models, message types, cursor events, and agent progress payloads. docs/development.md documents the Go plus pnpm monorepo, an 85 percent Go coverage gate, Playwright end-to-end tests, and the absence of a dedicated SDK test target.
Observed
- License
- MIT
- Server stack
- Go server with an embedded Svelte web application and SQL migrations
- Interfaces
- Web, Electron desktop, CLI, TypeScript SDK, REST API, and WebSocket API
- Server packaging
- Release binaries, Docker image, source build, Linux deb and rpm packages
- Platform support
- Server releases for Linux, macOS, Windows, and FreeBSD; desktop clients for macOS, Windows, and Linux
- Storage
- SQLite and local uploads by default, with Postgres and Cloudflare R2 options
- Testing structure
- Go tests, Playwright end-to-end tests, and desktop tests are documented; the TypeScript SDK has no dedicated test target
Read from README.md, go.mod, package.json, packages/sdk-ts/src/index.ts, docs/cli.md, docs/sdk.md, docs/README.md, docs/fakeco.md, docs/desktop.md, docs/install.md, docs/releasing.md, docs/data-model.md, docs/deployment.md, docs/quickstart.md, docs/development.md.
What it can do
Send messages to other users
Text message → Delivered message in chat conversation
Communicate with AI agents
Text message or query → Agent response in chat thread
Create group conversations with agents
User selection of participants and agents → Multi-participant chat room with AI agents
Process OpenClaw protocol messages
OpenClaw-formatted data → Native messaging interface display
Manage agent permissions in conversations
Agent selection and permission settings → Configured agent access levels
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.