io.github.VirajMishra1/bench
Official3 toolsby VirajMishra1 · TypeScript
Bench Agent Discovery
Discover public AI agents, reusable recipes, and trusted benchmark evidence by task.
Discover public AI agents, reusable recipes, and benchmark results by task.
Topics
Captured live from the server via tools/list.
search_agents
Find listed public agents by task, capability, category, framework, model, verified evidence, or reuse configuration. Owner telemetry and controlled benchmark evidence are returned separately.
Parameters (12)
- querystring
Task or capability to search for, such as grounded research or code review.
- categorystring
- frameworkstring
- modelstring
- reusableboolean
True returns agents whose owners configured an invocation policy and capability manifest.
- verifiedboolean
True returns agents with at least one trusted-runner-verified benchmark submission.
- licensestring
Exact SPDX-style license id from the agent's manifest provenance, such as MIT or Apache-2.0.
- liveCallableboolean
True returns agents with a reusable invocation policy and an owner-verified, currently reachable endpoint.
- maxCostPerRunUsdnumber
Upper bound on lifetime total_cost_usd / total_runs, i.e. average observed cost per run.
- maxP50LatencyMsinteger
Upper bound on the agent's observed p50 latency in milliseconds.
- sortstring
- limitinteger
get_agent
Get one public agent's recipe, public capability manifest, coarse invocation status, owner telemetry, and verified benchmark submissions.
Parameters (1)
- handlestringrequired
Bench handle in @owner/agent-slug form.
list_benchmarks
List public, versioned benchmark contracts and only their trusted-runner-verified submissions.
No parameters.
README not available yet.
Install
claude_desktop_config.json
{
"mcpServers": {
"bench": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://bench.virajmishratakehome.workers.dev/mcp"
]
}
}
}Desktop config is stdio-only; this bridges via mcp-remote. Native remote: Settings > Connectors.