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

Annotations cover the safety profile (readOnly, non-destructive, openWorld, non-idempotent), and the description adds real value on top: BYOK auth requirement, the public-address-only fetch restriction, and the shape of the result (per-case PASS/FAIL, pass rate, overall verdict, latency). It does not explain cost/rate implications of calling an external provider, which is the main residual gap.

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 dense sentences, front-loaded with the core action and the key behavioral facts; nothing is filler. The second sentence is a long run-on chain of clauses, which slightly hurts scannability but costs no information.

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?

An output schema exists, so spelling out return values was optional, and the description covers auth, input constraints, and assertion semantics for a 3-required-of-8 parameter tool. Minor gaps remain around system_prompt's effect and base64 input handling, but an agent has enough to call it correctly.

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?

With only 50% schema coverage, the description compensates well by defining assertion semantics that the schema only enumerates — notably that semantic_contains means TF-IDF cosine ≥ 0.4 — and clarifying that the threshold governs the overall PASS. It adds little for system_prompt, image_base64, or image_mime_type, keeping it short of a 5.

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?

States a specific verb and resource: test a vision model (with named providers) on one image with question/assertion test cases. The 'one image' scope implicitly separates it from run_vlm_test_suite_batch, and the enumerated assertion types make the tool's job concrete.

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?

Gives clear operating context: BYOK, key must match the model's provider, and the image must be reachable at a public address. It never names an explicit alternative (e.g. run_vlm_test_suite_batch for multiple images, or prompt_test_suite for text-only), so routing is implied rather than stated.

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