metaharness_bench
ADR-153 supporting verb — create or verify bench suites used by metaharness_evolve --bench. Bench suites are JSON files of {input, expectedOutput, weight} tasks; scoring against a fixed corpus decouples evolution from flaky/slow/undersized
This record as markdown: /tools/io-github-ruvnet-claude-flow/metaharness-bench.md
What metaharness_bench does on Claude Flow
AI agents use metaharness_bench to create or update resources in Claude Flow, usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Claude Flow environment.
Why metaharness_bench is rated Medium
This tool creates or modifies benchmark suite JSON files, which are configuration data used to drive evolution processes. While it does not permanently destroy data (making it less severe than Destructive) or execute arbitrary code (Execute), it does reversibly create and modify structured data artifacts. The Write category applies because the tool's primary function is creating or updating benchmark suite files.
From the tool's definition The tool description states it can 'create or verify bench suites used by metaharness_evolve --bench', with bench suites being 'JSON files of {input, expectedOutput, weight} tasks'.
Attacks that exploit this kind of access
The rule that runs metaharness_bench safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Claude Flow, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For metaharness_bench, this is the rule to start with:
metaharness_bench 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.
The button opens the PolicyLayer dashboard: create your workspace, connect Claude Flow, apply this rule, and every metaharness_bench call is checked against it from then on.
Questions about metaharness_bench
ADR-153 supporting verb — create or verify bench suites used by metaharness_evolve --bench. Bench suites are JSON files of {input, expectedOutput, weight} tasks; scoring against a fixed corpus decouples evolution from flaky/slow/undersized. It is categorised as a Write tool in the Claude Flow MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Claude Flow MCP server in PolicyLayer and add a rule for metaharness_bench: 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 Claude Flow. Nothing to install.
metaharness_bench is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the metaharness_bench 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 metaharness_bench. 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.
metaharness_bench is provided by the Claude Flow MCP server (claude-flow). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Claude Flow, and thousands of servers like it.
This server
Across the catalogue