feedback · Unfragile MCP: Risk & Policy

WHAT IT DOES

What feedback does on Unfragile

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

WHY IT NEEDS A POLICY

Why feedback needs a policy

This tool writes feedback data back to the system to influence future recommendations. It creates/updates records in the learning system, making it a Write operation. It is reversible in the sense that individual feedback entries could be corrected, and the blast radius is low since misuse would only skew recommendations slightly rather than causing destructive or financial harm.

From the tool's definition 'Report whether a recommended tool worked or not' and 'the Unfragile graph uses this feedback to improve future recommendations' — submits feedback data that modifies the recommendation graph

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

feedback Rate-limited

feedback 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": {
    "feedback": {
      "limits": [
        {
          "counter": "feedback_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 Unfragile, so you can run agents against it safely.

FAQ

Questions about feedback

What does the feedback tool do? +

Report whether a recommended tool worked or not. This closes the learning loop — the Unfragile graph uses this feedback to improve future recommendations. Call this after trying a tool from search results. It is categorised as a Write tool in the Unfragile MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.

How do I enforce a policy on feedback? +

Register the Unfragile MCP server in PolicyLayer and add a rule for feedback: 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 Unfragile. Nothing to install.

What risk level is feedback? +

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

Can I rate-limit feedback? +

Yes. Add a rate_limit block to the feedback 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 feedback completely? +

Set action: deny in the PolicyLayer policy for feedback. 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 feedback? +

feedback is provided by the Unfragile MCP server (savirinc/unfragile-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.