CoworkerCAUTION
Open-source AI agent with MCP UI, app builder, A2A protocol, skills marketplace, and multi-provider chat. Built with Mastra. Alternative to OpenClaw.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Open-source AI agent built with Mastra. Acts as an AI team member that handles tasks, answers questions, and manages workflows via chat. Supports OpenAI, Anthropic, Google Gemini, NVIDIA, Groq, Kimi, and any OpenAI-compatible provider.
Features
- AI chat assistant — answer questions, draft content, summarize documents, and brainstorm ideas
- App builder — Lovable-like builder for creating internal dashboards and tools, maintained by agents with git version control
- MCP UI — visual interface for managing MCP servers and building agent-maintained internal dashboards
- MCP registry — discover and install MCP servers from the built-in registry
- Skills marketplace — install community-built skills from ClawHub and skills.sh
- A2A protocol — let other AI agents discover and communicate with Coworker via Agent-to-Agent protocol
- Scheduled tasks — run recurring AI workflows on a cron schedule via Inngest
- Google Workspace — manage emails, calendar, and docs through natural language
- WhatsApp bridge — interact with your AI agent via WhatsApp messages
- File workspace — upload, manage, and reference files in agent conversations
- API access — secure API key authentication for programmatic access and integrations
- Multi-provider AI — switch between OpenAI, Anthropic, Google, NVIDIA, Groq, Kimi, or any OpenAI-compatible endpoint
Screenshots
083cd6e4b776OBSERVED · 2026-10-08Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add coworker --env COWORKER_API_TOKEN=${COWORKER_API_TOKEN} --env GOG_GOOGLE_CLIENT_SECRET=${GOG_GOOGLE_CLIENT_SECRET} --env GOG_KEYRING_PASSWORD=${GOG_KEYRING_PASSWORD} -- npx -y [email protected]{
"mcpServers": {
"coworker": {
"command": "npx",
"args": [
"-y",
"[email protected]"
],
"env": {
"COWORKER_API_TOKEN": "${COWORKER_API_TOKEN}",
"GOG_GOOGLE_CLIENT_SECRET": "${GOG_GOOGLE_CLIENT_SECRET}",
"GOG_KEYRING_PASSWORD": "${GOG_KEYRING_PASSWORD}"
}
}
}
}Exposed tools (4)
2 read · 2 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
Build | write | Execute tasks and write code |
Coworker | read | An AI team member that helps with tasks, answers questions, and manages workflows. |
Fast | read | Quick responses |
Plan | write | Research and create plans |
Trust audit
CAUTIONgrade B · trust 83/100 Install with care. The audit found things worth knowing before you trust its output.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | WARN |
| L1 | Static analysis of the code | WARN |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (10 observation(s))
- Network
- declared (9 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- found
Findings (18)
icon.icns
DATABASE_URL=postgresql://postgres:password@localhost:5432/mastra
"@whiskeysockets/baileys": ["@whiskeysockets/[email protected]", "", { "dependencies": { "@cacheable/node-cache": "^1.4.0", "@hapi/boom": "^9.1.3", "async-mutex": "^0.5.0", "libsignal": "git+https://
} from '../../mastra-client'
import { useAppStore } from '../../stores/useAppStore'import { useAppStore } from '../../stores/useAppStore'import { MASTRA_BASE_URL, setMastraBaseUrl, setMastraApiToken } from '../../mastra-client'import type { McpRegistryItem } from '../../mastra-client'? new Blob([Uint8Array.from(atob(content), c => c.charCodeAt(0))])
const binary = atob(content)
data: Uint8Array.from(atob(content), (c) => c.charCodeAt(0)),
const binary = atob(content)
const binary = atob(content)
@ai-sdk/react, @codemirror/lang-css, @codemirror/lang-html, @codemirror/lang-javascript, @codemirror/lang-json, @codemirror/lang-markdown, @codemirror/lang-python, @electron-toolkit/preload
@mastra/ai-sdk, @mastra/core, @mastra/e2b, @mastra/fastembed, @mastra/libsql, @mastra/loggers, @mastra/mcp, @mastra/memory
2. Load environment variables (for Node.js):
- **API access** — secure API key authentication for programmatic access and integrations
coworker.pen
Gates applied: no_behavioural_pass.
083cd6e4b776full audit observations/trust-audit/mcp-server/array-ventures__coworker.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-08 | 083cd6e4b776 | CAUTION | B | 83 | first audit |
Questions
What is the Coworker MCP server?
Open-source AI agent with MCP UI, app builder, A2A protocol, skills marketplace, and multi-provider chat. Built with Mastra. Alternative to OpenClaw.
What tools does Coworker expose?
4 in total: 2 read-only, 2 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is Coworker safe to connect to an agent?
With care. The audit graded it B (83/100) and found 18 things worth knowing before you trust this server, listed below with the exact line each was found on.
What credentials does Coworker need?
It reads COWORKER_API_TOKEN, GOG_GOOGLE_CLIENT_SECRET and GOG_KEYRING_PASSWORD from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does Coworker run?
It speaks streamable-http, so it runs as a service you connect to over the network. It is published on npm as coworker at 0.2.0.
How current is this page?
The grade is for one exact copy of the source (083cd6e4b776), read on 2026-10-08. The repository is watched and re-audited when it changes.