Start Until Unanimous

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

Annotations are thin (readOnly=false, idempotent=false, destructive=false), so the description carries real weight and delivers: background execution with concrete stop conditions, evidence that is 'persisted durably and transported automatically,' and version-pinned semantics referencing run_until_unanimous. It stops short of warning that 'real API' rounds incur real cost or telling the caller how results are later retrieved.

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 sentences with the purpose and stop conditions front-loaded in the first sentence. Sentence 2 partially duplicates the evidence schema description, and sentence 3 defers substance via 'see that tool for details' rather than explaining it, but there is no outright waste.

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?

A 15-parameter background-job starter with no output schema needs to state what it returns and how to monitor or cancel the job; the description mentions neither, forcing the agent to infer session_poll and session_cancel_job from sibling names. It is adequate for initiating the job but incomplete for the full lifecycle.

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 33% schema description coverage, the description should compensate for undocumented parameters, but it only addresses `evidence` (which the schema already describes in greater detail) and `caller` (deferred to run_until_unanimous). Ambiguous enums like mode ('ship' vs 'circular') and the roles of lead_peer, until_stopped, and max_cost_usd are left 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 opens with a specific action — 'Start real API generation/revision rounds in the background' — and names the three termination conditions (unanimity, max_rounds, budget limit). The 'background' qualifier hints at the distinction from the synchronous run_until_unanimous, but it never names session_start_round as the single-round alternative, so sibling differentiation is incomplete.

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 positions this as the background variant and defers `caller` + relator-lottery semantics to run_until_unanimous, and the evidence parameter adds an explicit exclusion: 'Do not call session_attach_evidence for this routine path.' However, the tool never states when to prefer this over session_start_round or run_until_unanimous, leaving the when-to-use matrix partial.

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