Iterate on an existing search session

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

With no annotations, the description carries the full burden. It discloses dedup against previously shown images and that the LLM interprets feedback into keyword changes via refine_search. However, it doesn't mention side effects like session mutation, reversibility, or any potential rate limits. It covers some key behaviors but not all.

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?

Two sentences with zero waste. The primary usage is front-loaded, and the key behaviors (dedup, LLM interpretation) are clearly stated. Very efficient.

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 7 parameters, no output schema, and no annotations, the description is relatively brief. It covers core mechanics but omits details on several parameters and doesn't describe the return value or any side effects. It's adequate but not fully complete for complex usage.

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 coverage is only 43%, so the description must compensate for undocumented parameters like count, criteria, safe_search, and session_id. It adds meaning for feedback (referencing cell ids) and keywords (explicit replacement) but does not explain the remaining parameters. The description fails to bridge the coverage gap for those.

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's purpose: it iterates on an existing search session by taking feedback referencing round-qualified ids, producing the next round with dedup. It distinguishes from siblings by implying it's not for starting a new session (image_search_start) but for refinement, and the title reinforces 'existing'.

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?

It gives clear usage instructions: provide feedback with round-qualified ids and optionally explicit keywords. It implies when to use (with an existing session) but does not explicitly mention when not to use or name alternatives. Context from siblings suggests this is for continuation, but the description alone doesn't state exclusions.

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