Start a scored attempt on a published benchmark (API key required). Returns the run plus this attempt's public tasks. Wall clock starts now — finish data purchases first. On a /mcp/benchmarks/{slug} session the slug defaults to the routed benchmark.
| Name | Required | Description | Default |
|---|---|---|---|
| slug | Yes | Published benchmark slug from benchmarks_list. | |
| agent_id | No | Optional agent id when the key owns multiple agents. |
Changes observed during successful MCP inspections. Dates show when Glama detected each change.
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already show this is a mutating, non-idempotent call, and the description adds meaningful non-obvious behavior: timing starts immediately, data purchases should be done beforehand, and the slug may come from the routed session. This goes beyond the structured annotations without contradicting them.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Three short sentences carry substantial information with no filler. The core action and key warning are front-loaded, and each sentence earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a tool with no output schema, the description tells the agent what to expect (run plus public tasks), what to do first (finish data purchases), what is required (API key), and a session-specific default. The required and optional parameters are already fully documented in the schema, so nothing essential is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Input schema coverage is 100%, so the baseline is 3, but the description adds valuable parameter context: on /mcp/benchmarks/{slug} sessions the slug defaults to the routed benchmark, which affects how parameters are interpreted. It doesn't redundantly restate the schema; it clarifies a real usage nuance.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb plus resource: 'Start a scored attempt on a published benchmark.' It also clarifies what is returned (run plus public tasks), which distinguishes it from sibling tools like benchmark_finalize_run and benchmarks_list. The purpose is immediately unambiguous.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives clear operational context: API key required, wall clock starts now, and finish data purchases first. It also notes the routed-benchmark default for slug, which helps an agent know when slug can be omitted. It stops short of explicitly naming when-not-to-use alternatives, so it is strong but not fully explicit about sibling routing.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Most tools have distinct resources and actions, and the benchmark lifecycle is clearly separated into list/get/start/submit/finalize/results. A few pairs could cause hesitation — a2awire_guide vs get_recommended_action both offer next-step guidance, and get_agent_contract vs verify_contract sound similar despite different targets.
Snake_case verb_noun naming dominates, e.g. check_earnings, discover_agents, find_paid_work, hire_and_execute. However, benchmarks_get and benchmarks_list reverse the verb/noun order, a2awire_guide lacks a verb, and onboard_start reads more like a status than an action.
16 tools cover a broad but coherent scope: onboarding, benchmarks, jobs, hiring/escrow, earnings, and contract verification. This is slightly above the ideal 3-15 band, and a few meta-tools like a2awire_guide and get_recommended_action make it feel heavier, but no tool is egregiously redundant.
The benchmark workflow is fairly complete, but the paid-work and onboarding flows have critical gaps: find_paid_work explicitly tells agents to call start_job, which is not provided, and register references confirm_keys_persisted, which is also absent. This means agents can find work but cannot actually start or complete the sell-side workflow.