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

Annotations only declare read-only/idempotent/non-destructive; the description goes well beyond them by disclosing the access model: unpaid lists return the first 5 full rows plus locked.count/locked.by_type, records expose only the subject and first 3 related names per section, and contact values and decision-maker names are masked to types and counts. It also tells the agent that every answer self-reports what it withheld via `entitlement` and `locked`, which is exactly the behavioral context the schema and annotations cannot supply.

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-loads the return shape before filters, alternatives and access rules, so the most important information lands first, and sentences are dense rather than padded. The closing plan/upsell sentence ('7 days free at https://...') is arguably promotional rather than definitional, which keeps it from being a clean 5.

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?

With 22 params, no output schema and only 55% schema description coverage, the description carries the burden well: it describes the returned card contents, the ID format, the filtering vocabulary, the sibling alternative, and the full entitlement/truncation model. An agent can predict both the request contract and the response shape without additional fields.

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?

At 55% schema coverage across 22 params, the description compensates by mapping filter phrases to parameters: 'states or region' (state/states/region), 'canonical industries' (sector/industries), 'direct investing' (invests_directly), 'manager allocation' (allocates_to_managers), 'recent activity' (recent_activity_days), 'AUM' (min_aum_usd), 'an event type since a date' (has_event_type/event_since) and 'a named co-investor' (co_invested_with). It adds the sorted-by-one-column semantics for `sort` too, but leaves query, city, counterparty_kind, limit and cursor untouched.

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 ('Family offices as compact cards') and enumerates the returned fields (class with confidence, direct investing, sectors, check size, activity, relationship counts) plus the ID namespace (dfx:fo:). It also names the sibling it is NOT ('for a ranked answer to a buyer's criteria use rank_family_offices'), so an agent can separate it from rank_family_offices and search_family_office_investments 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?

Explicitly routes the ranked-answer use case to rank_family_offices, and enumerates the filter axes that this tool is for (states/region, industries, direct investing, manager allocation, activity, AUM, event-since, co-investor). Context is clear, but it never states when to prefer get_family_office for a single record or search_family_office_investments for deal-level data, so a full when/when-not matrix is not present.

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