Propose this week's candidates

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

Annotations already mark this as non-read-only, but the description adds genuine behavioral context beyond them: it stores state, discloses the replace-on-recall semantics ('Re-calling for the same week_start replaces the candidates'), and describes the return payload (numbered candidates, ballot_text with {vote_link} placeholder, per-member vote links). No contradiction with destructiveHint=false, since replacing its own prior proposal is disclosed normal overwrite behavior rather than hidden destruction.

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 sentences, action front-loaded ('Start a weekly cycle'), followed by outputs and the key caveat. Every clause earns its place; no fluff, no restating of the title.

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?

For a state-changing tool with an output schema, coverage is strong: return structure is explained despite the schema already documenting it, cardinality is set, and the destructive edge case (replacement) is disclosed. Minor gaps: no explicit pointer to message_group as the dependent next step (only the placeholder hints at it), and no mention of whether replacing candidates affects already-cast votes.

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 schema coverage at 50%, the description partially compensates: 'normally 10' informs the cardinality expected for the recipes array, and the week_start replacement behavior adds lifecycle semantics to that parameter. Several schema-level gaps remain (title, servings, source_url, tags, steps at the top-level recipes property), and the description does nothing to fill those, though the schema's strongest descriptions (image_url, ingredients) already cover the highest-risk fields.

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 verb+resource+outcome: 'Start a weekly cycle: store the candidate recipes... and return numbered candidates, a ballot_text..., and each member's vote link.' This fully differentiates it from siblings like add_candidates (which appends rather than starts) and set_this_weeks_recipes (which finalizes rather than proposes).

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?

'Start a weekly cycle' gives implied context that this is the entry point for a new voting round, and the replacement caveat implies re-calling refreshes instead of erroring. However, it never explicitly names alternatives or says when not to use it — notably, the near-sibling add_candidates exists in the same domain and no distinction is drawn between 'propose a fresh cycle' and 'add to an existing one.'

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