Agent BoardSAFE
Open-source multi-agent task board for OpenClaw. Kanban + DAG dependencies + MCP server + auto-retry + audit trail. Built for autonomous AI agent teams.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
🎯 Agent Board
Multi-agent task orchestration for OpenClaw and AI agent teams. Kanban dashboard · REST API · MCP server · DAG dependencies · Auto-retry · Audit trail
Features • Quick Start • OpenClaw Integration • API • MCP • Dashboard • Architecture
Why Agent Board?
Running multiple AI agents without coordination is chaos. Each agent works in isolation, tasks get duplicated, failures go unnoticed, and there's no way to build multi-step workflows.
Agent Board fixes this. It's a task management system purpose-built for AI agent teams — whether you're running OpenClaw agents, Claude, or any LLM-based agents.
- Agents pick up work from the board via heartbeat polling or webhook notifications
- Dependencies are enforced — Agent B can't start until Agent A finishes
- Failed tasks auto-retry — no human intervention for transient failures
- Task chaining builds pipelines — when one agent finishes, the next one starts automatically
- Full audit trail — know exactly who did what, when, and why
- MCP native — agents interact thro
0e40f1de6344OBSERVED · 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 agent-board --env AGENTBOARD_API_KEYS=${AGENTBOARD_API_KEYS} --env AGENTBOARD_WEBHOOK_SECRET=${AGENTBOARD_WEBHOOK_SECRET} --env OPENCLAW_HOOK_TOKEN=${OPENCLAW_HOOK_TOKEN} -- npx -y [email protected]{
"mcpServers": {
"agent-board": {
"command": "npx",
"args": [
"-y",
"[email protected]"
],
"env": {
"AGENTBOARD_API_KEYS": "${AGENTBOARD_API_KEYS}",
"AGENTBOARD_WEBHOOK_SECRET": "${AGENTBOARD_WEBHOOK_SECRET}",
"OPENCLAW_HOOK_TOKEN": "${OPENCLAW_HOOK_TOKEN}"
}
}
}
}Exposed tools (14)
6 read · 6 write · 2 destructive. Blast radius: 2 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
board_add_comment | write | Add a comment to a task |
board_create_project | write | Create a new project |
board_create_task | write | Create a new task in a project |
board_delete_project | destructive | Delete a project and all its tasks |
board_delete_task | destructive | Delete a task by ID |
board_get_project | read | Get project details with its tasks |
board_get_task_thread | read | Get task summary with all comments (for agent context) |
board_list_comments | read | List all comments for a task |
board_list_projects | read | List all projects with optional filters |
board_list_tasks | read | List tasks with optional filters |
board_move_task | write | Move a task to a different column (backlog, todo, doing, review, done, failed) |
board_my_tasks | read | List tasks assigned to a specific agent |
board_update_project | write | Update project fields |
board_update_task | write | Update task fields (status, assignee, priority, etc.) |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (3 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (2)
board_delete_project, board_delete_task
@modelcontextprotocol/sdk, express, zod, @types/express, @types/node, @types/supertest, supertest, typescript
Gates applied: no_behavioural_pass.
0e40f1de6344full audit observations/trust-audit/mcp-server/quentintou__agent-board.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 | 0e40f1de6344 | SAFE | B | 89 | first audit |
Questions
What is the Agent Board MCP server?
Open-source multi-agent task board for OpenClaw. Kanban + DAG dependencies + MCP server + auto-retry + audit trail. Built for autonomous AI agent teams.
What tools does Agent Board expose?
14 in total: 6 read-only, 6 that write, and 2 that can delete or overwrite (board_delete_project, board_delete_task). Every one is listed on this page with its risk.
Is Agent Board safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 2 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does Agent Board need?
It reads AGENTBOARD_API_KEYS, AGENTBOARD_WEBHOOK_SECRET and OPENCLAW_HOOK_TOKEN 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 Agent Board run?
It speaks stdio, so it runs as a local process your client starts. It is published on npm as agent-board at 0.1.0.
How current is this page?
The grade is for one exact copy of the source (0e40f1de6344), read on 2026-10-08. The repository is watched and re-audited when it changes.