verify_citations

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

Beyond the readOnlyHint annotation, the description discloses 'ZERO LLM calls', the three output buckets and their meanings, that 'unverifiable' is 'reported separately and NEVER as fabrication', and that for Typst it also reports broken references. This is rich behavioral context that guides the agent's expectations and safety.

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?

While several sentences long, every clause earns its place by explaining a distinct behavioral aspect or parameter nuance. The main purpose is front-loaded, and the use-case guidance is at the end. No filler or redundancy.

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 tool with no output schema, the description fully specifies what the agent will receive (three buckets with meanings), the input details, and edge cases (no identifier, coverage gap, book, rate-limit). It also notes Typst-specific broken references. Everything needed to call and interpret the result is present.

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 already has 100% coverage, and the description greatly enriches both parameters: for bib_path it explains auto-discovery logic (#bibliography(...), otherwise every .bib); for project_id it clarifies what it is not (not a path, not '.', not a name/title), how to obtain it (list_projects or URL), and the remote vs. local context. This far exceeds the schema's basic descriptions.

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 clear, specific verb phrase 'Detect likely-hallucinated references' and specifies the resource ('the project's bibliography'), then details the exact behaviors (BibTeX/Hayagriva, DOI/arXiv check, three buckets). This is unambiguous and distinct from the listed siblings, which are file/compile/project operations.

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?

Explicitly states when to use ('Use to sanity-check an AI-assisted draft before submission'), when not to (do not use for general local file I/O; prefer read/grep for local checkout), and how to obtain the project_id. Also instructs to call list_projects first when unsure, giving a clear decision tree.

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