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

Annotations only give the safety/consistency triad (readOnlyHint=false, idempotentHint=true, destructiveHint=false). The description adds substantive behavioral context beyond that: drafts are pending and user-owned, they cannot approve/publish/modify public cards, raw transcripts and historical-conversation data are never accepted, and consent is a precondition rather than a parameter.

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?

Front-loaded with the core action and consent precondition, then the eligibility and prohibition clauses. Dense but almost every clause adds a real constraint; the ineligible-type enumeration is long but functionally load-bearing.

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?

Output schema exists, so return values need no explanation, and the description covers eligibility, scope assertion, consent, and provenance limits for a complex tool. It leaves parameter-level mechanics (idempotency via request_id, provider enum, relationship constraints) to the schema, which is a modest gap given 0% schema description coverage.

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% on a deeply nested four-parameter schema, so the description carries the burden but only partly delivers: it locks knowledge_scope to general_knowledge and implies provenance must be current-conversation, yet says nothing about request_id, provider, or the cards array fields (title, summary, relations, evidence_selectors). Most parameter meaning must still be inferred from the JSON Schema itself.

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 (create card drafts) with scope qualifiers (pending, user-owned, general-knowledge) and distinguishes itself from the sibling create_knowledge_bundle_drafts by naming cards rather than bundles. An agent can tell what this creates without opening the schema.

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?

Explicit when-to-use (only after clear affirmative consent to specifically named, independently teachable knowledge in the current conversation) plus an explicit when-not list of ineligible content types and a disambiguation rule ('an assistant suggestion alone is not consent'). Hard exclusions for raw transcripts and public-card mutation are also stated.

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