record_observation · State Trace MCP: Risk & Policy

WHAT IT DOES

What record_observation does on State Trace

AI agents use record_observation to create or update resources in State Trace — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your State Trace environment.

WHY IT NEEDS A POLICY

Why record_observation needs a policy

This tool writes/stores observation data into the agent's working memory graph. It creates new memory entries but does not delete data, execute commands, or have financial implications. The blast radius is low as misuse would result in incorrect or polluted memory state, which is reversible in principle.

From the tool's definition "Record an observation from a tool or command the agent just ran"

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

record_observation Rate-limited

record_observation stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.

View as policy code

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "record_observation": {
      "limits": [
        {
          "counter": "record_observation_rate",
          "window": "minute",
          "max": 30,
          "scope": "grant"
        }
      ]
    }
  }
}

LIMIT THIS TOOL →

Instant setup, no code required.

PolicyLayer is an MCP gateway: every tool call is checked against your rule before it reaches State Trace, so you can run agents against it safely.

FAQ

Questions about record_observation

What does the record_observation tool do? +

Record an observation from a tool or command the agent just ran. It is categorised as a Write tool in the State Trace MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.

How do I enforce a policy on record_observation? +

Register the State Trace MCP server in PolicyLayer and add a rule for record_observation: 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 State Trace. Nothing to install.

What risk level is record_observation? +

record_observation is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.

Can I rate-limit record_observation? +

Yes. Add a rate_limit block to the record_observation 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.

How do I block record_observation completely? +

Set action: deny in the PolicyLayer policy for record_observation. 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.

What MCP server provides record_observation? +

record_observation is provided by the State Trace MCP server (agent-pattern-labs/state-trace). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.