TurboQuant rerank candidates

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

Annotations already declare readOnlyHint=true and openWorldHint=false, so safety is covered; the description then adds genuinely new behavioral context: the two-phase execution model, the pg_turboquant extension dependency, and the halfvec overload triggered by half_precision. It does not mention cost, latency, or what happens when candidate_limit < final_limit, but the additions go meaningfully beyond the structured fields.

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?

Three compact sentences, front-loaded with the operation and pipeline, then returns, then the precision switch and prerequisite. No filler, though the backtick-heavy notation (``final_limit``) and the trailing dependency sentence are slightly awkward in ordering.

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 the brief return-value note suffices, and the extension prerequisite is a valuable addition. But for a 12-parameter, 8-required tool with near-zero schema coverage, the description leaves most parameters — including the query_vector format and the metric/probes knobs — undocumented, which is a real gap for correct invocation.

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 8% — only `database` is documented in the schema — so the description carries the burden for 12 parameters. It clarifies final_limit (exact rerank target), half_precision, and ties candidate_limit to the approximate stage only via the phrase 'approximate retrieval', but leaves metric, probes, oversample_factor, id_column, embedding_column, query_vector (array vs string form), schema, and table entirely unexplained.

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 names a specific operation and resource: run tq_rerank_candidates against a turboquant index, with the two-stage pipeline (approximate retrieval then SQL-side exact rerank to final_limit) spelled out. That mechanism implicitly separates it from turboquant_approx_candidates and from generic vector_search, though no sibling is named explicitly.

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 a prerequisite (requires the pg_turboquant extension) and a mode switch (half_precision=True uses the halfvec overload), which helps an agent decide it can call the tool at all. However, there is no explicit guidance on when to choose this over turboquant_approx_candidates, vector_search, or the analyze_reranker_* tools, so the selection logic is left to inference.

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