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

The description contradicts the annotations: it says webhook and other asynchronous blocks 'really call the endpoint or the model' (the schema even cites stripe__refund as a sample tool), which is external-world interaction and real side effects, while the annotations declare openWorldHint=false and readOnlyHint=true. The disclosure is otherwise rich and safety-relevant, but it directly conflicts with two declared hints, so the contradiction penalty applies.

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 outcome and the dry-run guarantee are front-loaded, and the sentences are dense but each carries distinct information — the return shape has to be described in prose because there is no output schema. Some sentences (branch/delegate enter vs next vs end_of_branch) are long and could be split, but nothing is disposable.

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 correctly explains what the report contains (outcome, deciding block, reason, escalation path, per-step block/pipeline attribution). It also covers the nested sample/blocks parameters and the async side-effect caveat; only auth/permission expectations and any rate-limit behavior are left unstated.

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%, so the schema already documents all five parameters. The description adds cross-tool semantics (blocks 'same shape update_pipeline takes', default_* 'as update_pipeline takes it') and notes that delegate blocks expand to their targets' active revisions, but the per-parameter meaning is largely schema-borne.

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 — run a sample request through a set of blocks — and enumerates the exact outcome vocabulary (approve, deny, human, next), the deciding block, its reason, and the escalation path. This clearly separates it from get_pipeline (read config), update_pipeline (write config), and test_webhook_endpoint (single-endpoint test).

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 establishes the dry-run context ('Nothing is saved and no approval request is created') and tells the caller to pass the current blocks from get_pipeline to test what is active, cross-referencing update_pipeline's block shape. It never names an alternative sibling to use when a real request is wanted, but the when-to-use framing is unambiguous.

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