evaluate_retrieval

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full burden of behavioral disclosure. It usefully indicates that the operation runs on a test dataset and measures recall/precision, which conveys an offline evaluation context. However, it does not disclose the output format, whether any files are written, or whether this operation has side effects on configuration or logs.

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?

The description is well structured, front-loaded with the core purpose, and uses a scannable trigger list followed by parameter definitions. There is minor redundancy between 'For developers testing improvements' and 'This is a development/evaluation tool, not needed for normal usage,' but the content is otherwise economical.

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 simple two-parameter tool with no required parameters and defaults for both, the description provides enough information to invoke it: what it does, when to use it, and what the parameters mean. The main gap is the absence of any explicit statement about the return value or output format, which matters more because there is no output schema.

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?

The input schema provides only types and defaults, with 0% description coverage, so the included PARAMETERS section is essential and fully compensates. It clearly defines dataset_path as the path to the JSON test-query file and final_k as the number of results per query to evaluate. This gives the agent enough meaning to invoke the tool correctly.

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 first sentence clearly states a specific action and resource: 'Run retrieval evaluation metrics on a test dataset.' It also adds audience context ('For developers testing improvements') and explicitly positions the tool as a development/evaluation tool rather than normal usage. It does not explicitly differentiate from sibling tools like search_knowledge or evaluate_co_scientist_run, so it stops short of 5.

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 AUTOMATIC TRIGGERS section lists concrete conditions under which to call the tool, such as evaluating retrieval quality after configuration changes and measuring recall/precision. The line 'not needed for normal usage' provides a clear exclusion. It does not name an alternative tool to use instead, so guidance is good but not fully explicit.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.