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

Annotations already cover safety (readOnly, idempotent, non-destructive, closed-world), so the description's job is to add what they can't: it discloses the three possible statuses (clear|changed|attention), that output is bounded to at most 3 actions each with a source and falsifier, and that a new cursor is returned. It omits auth requirements and any rate/scope limits on how much of the stack can be sent.

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 value proposition and only four sentences, none wasted. The telegraphed clause 'New choice: decide. Live error: rescue.' and 'at most 3 actions (source, falsifier)' take a moment to parse, which is a small cost against otherwise tight structure.

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?

No output schema exists, and the description does describe the return contract (status set, bounded action list with provenance, new cursor), which is the main gap it had to close. For a 6-parameter tool with a nested policies object, the remaining omission is guidance on the policies fields and volume expectations.

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 83%, so the schema already carries parameter meaning; the description adds a real constraint not present there ('never a manifest') and clarifies that the cursor is the caller's last one, tying cursor to the returned value. It says nothing about the nested policies object, though its fields are largely self-describing.

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: a delta report of what changed for the caller's stack since the last call, enumerating the change classes (retired models, price changes, deprecated packages, pick changes). An agent can tell this apart from decide/rescue/depot without opening any 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 an explicit trigger ('at session start if your code calls third-party APIs'), an input constraint ('never a manifest'), and routes the two adjacent cases to siblings: 'New choice: decide. Live error: rescue.' Both alternatives and the conditions that select them are named.

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