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

With no annotations, the description carries the full burden of behavioral disclosure. It explains the probe methodology (daily, temperature 0, three repeats), the condition for reporting drift (agreement among repeats and disagreement with previous run, same-day disagreement is noise), and the limitation that the alias-to-model mapping cannot be answered retroactively without this record. This is rich behavioral context covering how the tool operates and its constraints, though it does not specify output format or error behavior.

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 concise at ~150 words but dense with essential information. It is well-structured: first a one-sentence summary, then methodology, then the drift detection rule, then view explanations. Every sentence adds value: no fluff, no repetition of schema content. The first sentence clearly states purpose, and the rest provides necessary behavioral context.

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?

Given the tool's complexity (5 optional parameters, 3 views, no output schema, no annotations), the description is quite complete. It explains the core methodology, the drift definition, the purpose of views, and the key constraint (no retroactive aliases). It lacks specification of return structure or error cases, but with no output schema, the agent might need more information about response format. However, for a data retrieval tool, this level of detail is strong and likely sufficient.

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?

Schema description coverage is 80% (4 of 5 params described: days, view, provider, severity; limit lacks description). The description adds meaning beyond the schema by explaining the 'aliases' view answers a question that cannot be answered retroactively, and that severity is only for events. It clarifies the purpose of each view in the context of drift detection, which adds value beyond the schema's simple labels. The one undocumented parameter (limit) is self-explanatory due to naming.

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 clearly states the tool records what changed behind a provider endpoint (alias resolution and behavior shifts) with specific details on methodology (daily fixed probe suite, temperature 0, three repeats). It distinguishes from siblings like model_archive and series_history by focusing on endpoint drift measurement, not just archival or historical series.

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 why this tool is needed (providers update models behind stable names) and when it is appropriate (to answer 'what was actually behind an alias on that day'). It implies usage for drift detection but does not explicitly state when not to use it or mention alternatives. The context of independent measurement suggests it complements rather than replaces other tools, but no direct exclusions or alternative names are given.

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