submit_answers

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

With no annotations, the description fully discloses tool behavior: the three enforced invariants (quote resolution, server-run absence search, algorithm-computed judgement), the requirement for override justification, the visibility of overrides in the report and ratification queue, and the distinction that incomplete calls require re-calling. It also warns that unratified items mean the assessment is not final. This is rich, honest context.

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 long (roughly 250 words) but every sentence carries operational weight, covering modes, answer format, enforcement, overrides, escalation, and finalization. It uses clear paragraph breaks and a numbered 'three things are enforced' list for readability. While it is more verbose than necessary for a simple tool, the complexity of the tool justifies the density; it loses a point only for being on the longer side.

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 (12 parameters, two modes, override and escalation logic) and the complete absence of schema descriptions or annotations, the description is remarkably thorough. It covers the answer item structure, the three enforcement mechanisms, the incomplete status and retry behavior, override justification and visibility, the finalization flow, and how to handle the returned report. Output schema exists, so the description doesn't need to detail return fields, but it still explains the meaning of the report and ratification queue.

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 0%, so the description must compensate. It explains the structure of `answers` in detail (fields per question, evidence-mode-specific requirements), clarifies `support` as the narrative, and defines the behavior of `override_judgement`, `override_justification`, `domain` (0 finalizes), and `overall_escalate`. However, it omits explicit explanations for parameters like `render`, `direction_of_bias`, `overall_override`, and `overall_direction_of_bias`, though these are less central and somewhat self-explanatory.

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 opens with a specific verb and resource: 'Submit one domain's signalling-question answers and get back the COMPUTED domain judgement, or (domain=0) finalize the assessment.' This clearly distinguishes the tool's two modes and sets it apart from siblings like render_report or assess_result. The resource is well-defined and the outcome is explicit.

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 provides extensive usage guidance: instructing to answer only questions the algorithm reaches, explaining the incomplete status and retry flow, detailing override and escalation requirements, and stating when to use domain=0 for finalization. It also clarifies what not to do (e.g., answering NA) and that the returned report is the artifact of record, not a custom table. This gives the agent clear decision rules for invoking the tool correctly.

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