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

With no annotations provided, the description fully carries the behavioral disclosure burden. It reveals side effects (emitting facts to the projection stream), idempotency ('duplicate fact_hashes skipped unless overwrite_existing=true'), return values ('facts_added + rules_matched + receipt_id'), and a server-side gating condition ('FACT_EXTRACTION_MODE'). This level of transparency exceeds typical descriptions and gives the agent crucial operational context.

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?

The description is information-dense but every sentence adds value: it covers the action, use case, idempotency, profiles, return values, and gating. There is no redundancy or filler, and the key facts are front-loaded—the opening sentence immediately establishes what the tool does. This is an exemplary balance of conciseness and completeness.

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?

The description is highly complete given the tool's complexity and the existence of an output schema. It explains when to use the tool, what it returns, the extraction profiles, and idempotency. Minor gaps remain: the meaning of 'seeded sessions' and how session_ids interacts with the scan scope is unclear, and the output schema (not provided in the prompt) would handle return value specifics. Still, the description provides sufficient context for most use cases.

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%, so the description must compensate for parameter meaning. It does explain extraction_profile by listing its enum values and their behavior, and overwrite_existing is clarified in the idempotency note. However, the session_ids parameter is not explicitly described; the phrase 'seeded sessions' is ambiguous about whether it refers to all seeded sessions or is filtered by session_ids. Since one of three parameters lacks clear semantics, the description only partially compensates for the missing schema documentation.

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 clearly states the tool's function: 'Scan the tenant's seeded sessions with rule-based extractors' and 'emit structured facts to the projection stream so they become queryable via enumerate_memory_facts.' It specifies the resource (sessions), the action (extract facts), and the outcome (queryable facts), effectively distinguishing it from sibling tools like enumerate_memory_facts, which is referenced as the query counterpart.

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 usage guidance is provided: 'Use when enumerate_memory_facts returns insufficient rows for aggregation, version-chain, or money questions and you suspect the fact exists but was under-predicated at ingest.' It also explains profile selection ('comprehensive' vs. narrower profiles) and emphasizes idempotency, making it safe to re-run. This clear guidance helps an agent decide when to invoke this tool versus alternatives.

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