// SCAN REPORT

RoadBoard

34 tools. 18 can modify or destroy data without limits.

1 destructive tool with no built-in limits. Policy required.

Last updated: 29 Jun 2026

18 can modify or destroy data

16 read-only

34 tools total

Community server · catalogue entry verified 29/06/2026

How to control RoadBoard ↓

TOOL MAP

What RoadBoard exposes to your agents

Read (16) Write / Execute (17) Destructive / Financial (1)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous RoadBoard tools

Destructive · High delete_task This tool irreversibly removes data with no recovery mechanism ('Irreversible', 'no undo'), matching the Destructive category definition. Write · Medium create_architecture_annotation The tool creates new annotation data attached to existing architecture nodes, which is a reversible write operation. Write · Medium create_architecture_edge This tool creates/modifies project architecture metadata by adding directed edges to a graph model. Write · Medium create_architecture_link This tool creates a new association/link between an architecture node and an existing entity (task, decision, milestone, or memory entry).

18 of RoadBoard's 34 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control RoadBoard

PolicyLayer is an MCP gateway — it sits between your AI agents and RoadBoard, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations

{
  "delete_task": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "create_architecture_annotation": {
    "limits": [
      {
        "counter": "create_architecture_annotation_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "get_architecture_map": {
    "limits": [
      {
        "counter": "get_architecture_map_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

Create a free account and register RoadBoard — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON ROADBOARD →

Instant setup, no code required.

FULL CATALOGUE

All 34 RoadBoard tools

DESTRUCTIVE 1 tools

Destructive delete_task Permanently delete a task. Irreversible — there is no soft-delete or undo. Use only when the task was created

WRITE 17 tools

Write create_architecture_annotation Attach a free-text annotation to an ArchitectureNode. Use for semantic context that cannot be derived from sou Write create_architecture_edge Create a directed edge between two ArchitectureNodes. Emit edgeType= Write create_architecture_link Link an ArchitectureNode to an existing RB entity (task, decision, milestone, memory entry). Use when the agen Write create_architecture_node Create an ArchitectureNode (workspace/module/service) in the project graph. Call once per workspace while pars Write create_architecture_repository Create a CodeRepository record for a project. Required before creating any ArchitectureNode. Typical agent flo Write create_decision Record an architectural or project decision Write create_handoff Create a structured handoff memory entry from current session activity. Use at end of session to preserve cont Write create_memory_entry Create a new memory entry in a project. Use Write create_phase Create a new phase (roadmap milestone) in a project. TITLE NAMING CONVENTION: Phase title MUST follow the for Write create_project Create a new project in RoadBoard. IMPORTANT: before calling this tool, if the user has not explicitly specifi Write create_task Create a new task in a project phase. If phaseId is omitted the first available phase is used automatically. I Write ingest_architecture One-shot orchestrator for agent-driven onboarding (B.2 flow). The agent scans the repository locally, builds a Write link_task_to_node Semantic wrapper over create_architecture_link: tie a task to one ArchitectureNode it will touch. Call this im Write update_decision Update an existing decision: change status, record outcome, add rationale, or mark as resolved. Write update_phase Update fields of an existing phase (title, status, dates, linked decision). Write update_task Update fields of an existing task (title, description, priority, phase, assignee, due date). Use update_task_s Write update_task_status Update the status of a task. When closing a task (status=done), provide a detailed completionReport describing

READ 16 tools

Read get_architecture_map Get the architecture graph for a project: all current nodes (apps, packages, modules) and edges (depends_on, i Read get_architecture_snapshot Get a compact architecture snapshot for a project: summary by node/edge type, top 5 highest-impact nodes, and Read get_node_context Get full context for an architecture node: annotations, links to tasks/decisions/memory, and impacted-by analy Read get_project Get details of a specific project Read get_project_changelog Generate a structured, agent-readable changelog for a project: task summary, active phases, recent decisions, Read get_project_memory List memory entries for a project Read get_user Look up a user by ID and return their public profile: id, username, displayName, email. Use this to resolve ra Read initial_instructions Call this tool ONCE at the start of every session. Returns the RoadBoard 2.0 MCP operational protocol: availab Read list_active_tasks List tasks for a project. Recommended for large projects: pass Read list_phases List phases for a project Read list_projects List all projects the current token has access to Read list_recent_decisions List decisions for a project, optionally filtered by status Read list_teams List teams the current user belongs to, each with its slug and the caller Read prepare_project_summary Generate a structured project snapshot for agent onboarding: project details, task counts by status, open task Read prepare_task_context Assemble full context for a task: project info, the task itself, sibling tasks in the same phase, and recent m Read search_memory Search memory entries for a project by keyword. Searches both title and body fields.

EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

FAQ

Questions about RoadBoard

Can an AI agent delete data through the RoadBoard MCP server? +

Yes. The RoadBoard server exposes 1 destructive tools including delete_task. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through RoadBoard? +

The RoadBoard server has 17 write tools including create_architecture_annotation, create_architecture_edge, create_architecture_link. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach RoadBoard.

How many tools does the RoadBoard MCP server expose? +

34 tools across 4 categories: Destructive, Execute, Read, Write. 16 are read-only. 18 can modify, create, or delete data.

How do I enforce a policy on RoadBoard? +

Register the RoadBoard MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every RoadBoard tool call.

Deterministic rules across all 34 RoadBoard tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON ROADBOARD →

Instant setup, no code required.

34 RoadBoard tools catalogued and risk-classified — across an index of 43,000+ MCP servers.