contact.search

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

With readOnlyHint already declaring safety, the description adds real behavioral context: per-source provenance rules (HubSpot + Salesforce only for `email`/`query`; `phone`/`company`/`alias` are local-warehouse only) and the fact that results carry `shape.pattern`. It stops short of explaining what patterns 1-4 actually mean, so the agent knows a signal exists but not how to read it.

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?

Three sentences, front-loaded with purpose, then sources, then differentiation — no filler. The middle sentence is dense with em-dash clauses and a bare `shape.pattern (1-4)` reference that is terse to the point of opacity, slightly hurting scanability.

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 7-param read tool with no output schema, it covers return fields (name, email, phone, company) plus provenance and the envelope discriminator, which is more than most. The one gap is that `shape.pattern` values are named but never decoded, leaving the agent to infer what 1-4 trigger.

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 the baseline is 3; the description earns above that by adding cross-parameter semantics the schema does not convey — mutual exclusivity of identifier kinds and which backends each identifier engages. This is meaning beyond field-level descriptions, not a repeat of them.

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 and resource ('Search the owner's contact graph') plus the identifier kinds it accepts, and explicitly contrasts itself with `memory.search` and the address-book-vs-past-recipe distinction. An agent can distinguish this from mail.search, deal.search, or enrichment.search without opening a 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 routing rule: 'Use this over `memory.search` when the user is asking about someone in their address book vs someone mentioned in a past recipe / audit row.' Also states the one-identifier-kind-per-call constraint, which is an operational precondition rather than just a parameter note.

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