AI Recommendation Check

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

Annotations only say this is non-read-only/open-world; the description adds the traits an agent actually needs: the run takes under a minute, it is asynchronous (returns check_id), saved results for the same business are reused for 7 days, and there is a limit of 3 buyer questions per website per day. It also discloses the privacy/data flow (email and business details go to Continuum, the founder may email personally) and that results appear only in chat with nothing auto-emailed.

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 trigger scenarios and then sequenced logically (what it does, timing, handoff tool, prerequisites, disclosure, exclusions, limits). It is long and repeats the email/privacy notice already carried by the schema, but nearly every sentence serves a distinct routing or invocation need.

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 no output schema, the description compensates fully: it explains what the check reports (whether each question named the business, who was named instead, one fix based on cited pages) and how the result is retrieved via get_ai_recommendation_result. Prerequisites, rate limits, privacy disclosure, and exclusions are all present, leaving no material gap for correct invocation.

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, but the description adds useful collection context: it tells the agent to ask the user for email, business name, website, customer-language category, and service area, and re-emphasizes that service_area is a fact about the business rather than the user's own location. That guides parameter sourcing in conversation beyond the raw field definitions.

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 precise verb+resource (runs a live check of whether ChatGPT/Perplexity recommend a given business) and enumerates the exact user phrasings it answers. It also distinguishes itself from its sibling by describing the handoff: this call starts the check and returns a check_id, while get_ai_recommendation_result fetches the outcome.

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?

Gives explicit when-to-use triggers (owner asking if AI recommends them, why competitors are recommended instead) plus a clear 'do not use this for' list covering SEO/Google ranking, marketing copy, person lookups, and general AI-search questions. It also names the required follow-up tool and the sequencing, so routing is unambiguous.

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

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

Annotations cover the safety profile (readOnly, idempotent, non-destructive, closed-world), and the description goes further with what annotations cannot express: the full status lifecycle (queued/running/done/failed/refused) and the actual return contract for ChatGPT and Perplexity results. The polling expectation is stated explicitly, which is exactly the behavioral context a caller needs.

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 core action and the sibling linkage before the status/return detail, and every sentence carries information. The status enumeration and result-content sentence are dense but not padded, though the description runs slightly long for a one-parameter tool.

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 no output schema, the description must describe what comes back, and it does: status values plus the per-engine breakdown (who named the business, who was named instead, one fix). Combined with the polling instruction, nothing an agent needs to call and interpret this tool is missing.

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 100% for the single check_id parameter, so the baseline is 3. The description only restates provenance ('using the check_id it returned'), adding no format or constraint detail beyond the schema's pattern and description.

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?

Names a specific verb+resource (get the result of a check) and ties it directly to the sibling that produces the input, check_ai_recommendation. The closing sentence explicitly rules out the sibling's job ('Do not use this to start a new check'), so an agent can separate the two tools without opening either 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?

Gives explicit when-to-use (retrieving a check's result by its check_id) and when-not-to-use (starting a new check). It also specifies the poll condition and cadence: if status is queued or running, tell the user and call again after ~15 seconds.

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