Findable

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

Annotations already declare readOnly, idempotent, non-destructive, non-open-world behavior, so safety is covered. The description adds genuine value by disclosing the completion payload (full result as text plus a Findable link), but says nothing about the pending state, intermediate status values, or error behavior for a bad id.

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?

Two short sentences with no filler, and the return-value information is front-loaded after the purpose. The opening clause is a verbless fragment, which costs a little clarity but not length.

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 carries the return-value burden; it covers the ready state but not the pending/in-progress state an agent polling this tool will most often encounter. For a simple one-param read tool it is nearly adequate, but the missing in-progress semantics leaves a real gap.

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?

There is a single required parameter and schema description coverage is 100% ("The id start returned."), so the schema already carries the semantics. The description adds nothing about the id's origin, format, or behavior for an invalid/unknown id, so baseline 3 applies.

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 identifies the resource (a Website SEO/AEO audit) and the operation (reading its current state, and the full result once ready). That is enough to separate it from start_audit, though the verb is expressed as a fragment ("Where ... stands") rather than an explicit 'check/poll status'.

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?

Usage is only implied: by describing a return that appears "when it is ready", it hints at polling after start_audit, but it never states when to call this versus start_audit, how often to poll, or what to do while the audit is pending.

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?

With readOnlyHint=false and openWorldHint=true already declared, the description goes well beyond them: it enumerates exactly what gets read (homepage, up to four internal pages, robots.txt, sitemap.xml, up to five competitors), discloses live OpenAI API web searches, cache-reuse rules with original dates, and that private client data is excluded. It also states the service will not draft or send outreach and returns an ID, which is unusually rich behavioral disclosure.

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?

It is one dense paragraph with no front-loaded summary or structure, and the trailing "Starts the work and returns an id. It usually takes several minutes." partly repeats the earlier audit-ID sentence. Most content is load-bearing given the absence of an output schema, but the wall-of-text form costs readability.

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 an 8-parameter tool with no output schema, the description covers what the audit does, what the report contains (sourced findings, target URLs, priorities, drafts, impact, review points, measurement steps), and the async return contract. It leaves out polling cadence and failure behavior, which an agent would still need.

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% and every parameter carries its own description, so the schema does the heavy lifting and a 3 is the baseline. The description reinforces semantics like competitor discovery and the no-connection constraint, but adds no format or value detail the schema lacks.

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 opening sentence gives a specific verb and resource ("Create a stored SEO/AEO audit of a public business website") and qualifies the scope ("without an account or data connection"). The closing "Returns an audit ID for retrieving progress" makes the async pairing with check_audit inferable, but the sibling is never named outright, so it falls short of full sibling differentiation.

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?

It states the operative context (no account or data connection required, optional provider snapshots, audit runs without them) and draws explicit boundaries ("cannot access other MCP connections, infer private metrics or edit websites"). It never explicitly says to poll with check_audit or when an agent should not start an audit, so exclusions are partial.

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