reject_all_candidates · OpenMemBrain MCP: Risk & Policy

WHAT IT DOES

What reject_all_candidates does on OpenMemBrain

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

WHY IT NEEDS A POLICY

Why reject_all_candidates needs a policy

This tool modifies candidate states in bulk, which is a write operation. While rejection could be considered somewhat destructive in intent, the candidates are 'pending' (not yet committed memory), and rejection is conceptually reversible — candidates could potentially be re-proposed or the action could be undone through system recovery.

From the tool's definition The tool 'reject_all_candidates' performs a batch operation to 'reject and remove all pending memory candidates' — it modifies the state of multiple candidates by changing them from pending to rejected status.

Attacks that exploit this kind of access

Recommended rule Ready

Verdict

reject_all_candidates Rate-limited

reject_all_candidates 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": {
    "reject_all_candidates": {
      "limits": [
        {
          "counter": "reject_all_candidates_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 OpenMemBrain, so you can run agents against it safely.

FAQ

Questions about reject_all_candidates

What does the reject_all_candidates tool do? +

Reject and remove all pending memory candidates at once. It is categorised as a Write tool in the OpenMemBrain MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.

How do I enforce a policy on reject_all_candidates? +

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

What risk level is reject_all_candidates? +

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

Can I rate-limit reject_all_candidates? +

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

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

reject_all_candidates is provided by the OpenMemBrain MCP server (mohamadalhusseinie/openmembrain). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.