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

Annotations only state readOnlyHint=false and idempotentHint=true, so the description carries the full burden of behavioral disclosure. It adds multiple non-obvious side effects: 'Updating project keys changes resolved settings on existing forms unless a form override is set,' 'Null Turnstile or webhook fields clear them,' and 'Missing fields render empty and never error.' It also warns against passing HTML, cautions that the files_url folder 'never put that URL in a public snippet,' and explains plan-based enforcement. There is no contradiction with annotations.

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 a single dense paragraph with several topical digressions—plan pricing, multipart POST file behavior, and inbox folder URLs—that are not strictly needed to invoke update_project. It is front-loaded with the core purpose, but later sentences drift into product-level behavior; a more structured layout with section breaks would make the guidance significantly easier to scan.

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?

Given 21 parameters, nested objects, plan restrictions, and no output schema, the description is remarkably comprehensive: it covers prerequisites, side effects, plan capacity rules, template token semantics, and a security warning. The only notable gap is the absence of explicit return-value or authorization information, but the surrounding context gives an agent enough to call the tool correctly.

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 lifts it by explaining nested email field structure, the template token language including fallbacks like '{{name | "New submission"}},' inheritance of upload settings ('forms inherit them'), and plan-restricted branding behavior not fully captured in the schema. It does not enumerate every property, but the schema already documents each, so the added semantics are meaningful rather than redundant.

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 opens with 'Use to patch website settings,' a specific verb and object that clearly identifies this as a mutation targeting project-level configuration. It further delineates scope with 'Do not use this to create forms' and explains that project keys affect existing forms unless overridden. However, it never explicitly contrasts itself with the sibling update_form, leaving the closest alternative differentiation implicit.

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?

The description gives explicit routing guidance: 'Use create_upgrade_link to hand a human a Stripe Checkout (or Customer Portal) URL' and warns 'Do not use this to create forms.' It also provides prerequisites such as 'Enable uploads on the project first' and plan-eligibility constraints like branding.mode being 'ignored on free except furrow.' It does not name update_form as the alternative for form-level overrides, though it notes project keys change resolved settings unless an override is set.

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