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

Goes well beyond the annotations by disclosing that suggestions are only staged and "never tracked" until the user Accepts via review_url, that the call consumes AI credits, and that the shortfall is reported but never filled. The run_key idempotency contract is spelled out in operational terms (same key on retry returns the staged result), which complements rather than repeats idempotentHint.

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 and the agreed-run precondition are front-loaded, and most sentences carry operational weight (credit spend, retry, staging, shortfall). The distribution sentence duplicates what the schema already says about branded/comparison/intent shares, making it the one line that could be trimmed.

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 9-parameter, credit-consuming, non-destructive mutation with no output schema, the description covers what an agent needs: consent gate, cost, staging/activation lifecycle, retry semantics, and the fact that shortfalls surface as reported reasons. Return behavior is described qualitatively (staged result, review_url), which compensates for the absent output schema.

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 100%, so a 3 is the baseline; the description adds genuine value by explaining the provenance of parameters — seed_source/seed_query_ids from read_seed_queries, topic_ids from confirm_topics or read_prompt_portfolio, prompt_set_id from read_prompt_portfolio — and clarifying that run_context is a one-off note never saved as a business fact. The distribution explanation largely restates the schema text, which is why it isn't a 5.

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?

States a specific verb + resource ("Run CiteLadder's prompt generator for the project") and anchors it to the UI equivalent ("the Prompt Builder's Generate step"), which clearly separates it from siblings like prepare_add_prompts, update_prompt_text, and read_prompt_portfolio. An agent knows this generates suggestions rather than mutating prompts directly.

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?

Gives a clear precondition ("once the user agreed to the run") plus a rate/usage rule ("call it once per agreed run") and retry guidance (resend the same run_key). It also routes the agent to the correct upstream sources for parameter values (read_seed_queries, confirm_topics, read_prompt_portfolio). It stops short of explicitly contrasting with the manual alternatives such as prepare_add_prompts.

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