cos_retrieval_enrichment_check

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

While annotations already declare readOnly, idempotent, and non-destructive behavior, the description adds crucial context: 'The underlying LLM enrichment path is intentionally a stub.' This reveals that the tool does not perform enrichment itself but only recommends, and it surfaces the cost-warning aspect, which are non-obvious behavioral traits beyond the annotations.

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 compact, with a clear opening purpose, a brief rationale, and formatted Args/Returns sections. Every sentence adds value, and the structure makes key information (purpose, stub behavior, parameter, output) immediately scannable.

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 one optional parameter and a simple recommendation output, the description covers all essential context: what it does, why it exists, what input it takes, and what it returns. The output schema is present, and the explicit Returns block further clarifies the response shape, including the optional cost_warning field.

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 only parameter, lookback_days, has no description in the schema (0% coverage). The description compensates fully by defining it as 'Window of retrieval quality data (default 14),' giving the parameter clear meaning and default behavior.

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 opens with a specific verb and resource: 'Recommend whether to enable contextual retrieval enrichment.' This clearly distinguishes it from sibling retrieval tools like cos_retrieval_quality or cos_retrieval_cite, which likely focus on measuring or citing retrieved content rather than making a cost-aware recommendation.

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 explains this tool is used to make a metric-driven decision before 'anyone pays the Haiku bill,' giving clear context for when to invoke it—as a gate before enabling enrichment. It does not explicitly name alternatives or exclusions, but the purpose and timing are unmistakable.

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