do_extract

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

No annotations are supplied, so the description carries the full behavioral burden. It does disclose the stateful two-phase pattern (Call 2 requires the extraction_id issued by Call 1) and enumerates rejection behavior ("one code per check") and jev_disagreements, which is useful. It says nothing about side effects, permissions, idempotency, or what happens if Call 1 succeeds but Call 2 is never made.

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?

Purpose is front-loaded and there is no filler, but the brace/bracket shorthand is hard to parse and forces the agent to decode notation rather than read prose. Terms like jev_disagreements and extractor_local appear without expansion, costing more in comprehension than is saved in length.

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 yet the description partially restates return values rather than adding interpretation. For a multi-call, nested-payload tool with zero schema coverage, it covers the field list adequately but omits how the two calls relate operationally (timeouts, retry, whether Call 2 is mandatory) and what the required 'extractor' identifier must be.

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 0% — the only declared parameter is an untyped 'payload' object with additionalProperties true — so the description is the sole source of field information. It compensates well by enumerating the payload keys for both calls and marking required fields with asterisks (extraction_id*, extractor*, records*). It still leaves opaque fields such as contributor_off_confirmed, extractor_local, and new_points 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 opening sentence names a specific verb and resource: "Extract decision labels from a one-turn transcript". It also discloses a non-obvious structural trait (two required calls), which helps an agent understand this is not a single-shot extract. It does not, however, differentiate the tool from near-name siblings like do_analyze or do_decide, and "decision labels" is left undefined.

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 implies usage through the call sequence — Call 1 first, then Call 2 with the returned extraction_id, plus a pass:2 audit variant. That is procedural guidance rather than selection guidance: nothing says when to reach for do_extract instead of do_analyze or do_decide, or what preconditions must hold before Call 1.

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