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

Annotations only declare readOnly=false, destructive=false, openWorld=true; the description goes well beyond by disclosing that secrets are minted server-side, working code is seeded, the operation is idempotent (re-running returns existing resource and skips existing seed files), and that turn_off preserves app data and owner Stripe setup. It also flags a user-facing confirmation dialog for app-connection, which the annotations cannot express.

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 core statement, then a clean bullet list of resources, and every section carries real information. It is dense and slightly uneven in ordering — the turn_off paragraph is wedged inside the resource list and app-connection trails after it — so structure could be tighter, but nothing is filler.

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 six-parameter, multi-mode mutation tool with no output schema, the description covers what gets created, what side effects occur, idempotency, ownership/permission constraints, and cross-tool references. An agent has everything needed to pick the right resource value and call it correctly on the first attempt.

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?

The schema leaves the 11-value resource enum undocumented; the description supplies per-value meaning for each, including which env vars get set and which files are injected. It also clarifies what action='turn_off' does versus the default 'provision', and explains app_target/app_functions/app_note beyond their terse schema strings (only editable apps, only functions in flootAppExports.tsx, note shown in dialog).

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 ('Provision a Floot-managed backend resource') and immediately enumerates all eleven resource kinds with what each creates. It also explicitly excludes adjacent concepts (email is not a resource; user's own key is not this tool), so an agent can tell provision_resource apart from request_external_resource and add_dependency 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?

Gives explicit routing rules and exclusions: SENDING email → @floot/email, user's own external key → request_external_resource, beta-gated enums unavailable, google-integration/payments overridden when the user wants their own client. The turn_off path adds preconditions ('Owner only; ask the user first and call it only once they agree') and the app-connection path adds an ordering constraint (write both apps, connect, publish both).

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