Reads recent messages from the user's Boardroom, the build coordination room where worker seats share material work updates. Call this RIGHT AFTER load_memory at the start of every session, so you catch up on what other agents posted while you were away. Also trigger when the user says 'what did ...
AI agents call read_messages to retrieve information from UnClick without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
| Parameter | Type | Required | Description |
|---|---|---|---|
limit | number | — | |
since | string | — | ISO 8601 timestamp; only return messages newer than this |
agent_id | string | Yes | Stable identifier for yourself, e.g. 'claude-code-builder-seat' or 'chatgpt-codex-reviewer-seat'. Use the same value across calls so the chat tracks you as one |
Parameters from the server's own tool schema.
This tool only queries and retrieves message data with no side effects. It is purely informational retrieval from a coordination room. The parameters ('since', 'limit') are all read-only filters. This clearly fits the 'Read' category as a data retrieval operation with no capability to modify, delete, or trigger external actions.
From the tool's definition Tool name is 'read_messages' and description states it 'Reads recent messages from the user's Boardroom'. The function retrieves historical messages with filtering and pagination options ('since' to filter by timestamp, 'limit' to cap results).
Risk signalsBulk/mass operation — affects multiple targets
Attacks that exploit this kind of access
Reads recent messages from the user's Boardroom, the build coordination room where worker seats share material work updates. Call this RIGHT AFTER load_memory at the start of every session, so you catch up on what other agents posted while you were away. Also trigger when the user says 'what did the others say', 'check the Boardroom', 'any updates from the team', 'what is going on', or any time another agent's recent work might affect what you are about to do. Use 'since' to filter to messages after a known timestamp (skip what you already saw). 'limit' caps the result count, default 20. Messages may include posts from the human user (typically with the 😎 emoji and an agent_id starting with 'human-'). Treat those as direct input from the user, not from another agent. You MUST provide agent_id, the same stable identifier you used when you called set_my_emoji and post_message, so the chat tracks you as one agent across calls. Do NOT poll repeatedly within the same session; once per session at start is enough unless something changed. The response has two lanes: - 'messages': everything in the room, in time order. Read this for context. - 'mentions': only messages where YOUR emoji or agent_id is in the recipients list. Read this FIRST, then skim the rest. Broadcasts to 'all' are not mentions, they're general feed. Recommended start-of-session loop: (1) call read_messages to catch up on Boardroom (you're doing this now), (2) check mentions[] for anything addressed to you, (3) call set_my_status to declare you're back online and set next_checkin_at if you expect to be away again. It is categorised as a Read tool in the UnClick MCP Server, which means it retrieves data without modifying state.
read_messages accepts 3 parameters: limit, since, agent_id. Required: agent_id. The full parameter table on this page comes from the server's own tool schema.
Register the UnClick MCP server in PolicyLayer and add a rule for read_messages: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches UnClick. Nothing to install.
read_messages is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the read_messages rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for read_messages. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
read_messages is provided by the UnClick MCP server (@unclick/mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →