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

With annotations only declaring the mutation/idempotency profile, the description adds genuinely new behavior: merge semantics (“Only the fields given change”), zero-as-deletion for limits, whole-policy removal via remove=true, and a required user confirmation. It omits permission/ownership requirements and the blast radius of changing a policy while actions are in flight.

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?

A single dense but front-loaded paragraph where nearly every clause carries a distinct field rule; nothing is redundant filler. It is compressed to the edge of readability in places (“the protection role”, “never” as a standalone gloss for blocked).

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 11-parameter mutation tool with an output schema (so return values needn't be explained), it covers merge behavior, deletion, defaults, and the confirmation step. The remaining gap is that owner/note are unaddressed and there is no cross-reference to overlapping sibling tools, which matters for a policy tool in a crowded policy surface.

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 0%, so the description carries the full burden, and it defines 9 of 11 parameters in concrete terms (per_action_usd “refused above”, approval_over_usd “a person approves above”, empty merchants = any, '*' default, 0 removes). It never explains owner or note, and its parentheticals are telegraphic, but it substantially compensates for the empty schema.

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?

Opens by defining the actual mandate (“What an agent may spend without a person”) and enumerates the specific knobs the policy controls, so the resource and effect are unmistakable. It never names a sibling it differs from, however, leaving the agent to distinguish it from set_budget, limits, or change_protection_settings on its own.

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 through field semantics; there is one procedural instruction (“Confirm with the user first”) but no statement of when to pick this over the many adjacent policy tools (set_budget, protection_settings, agent_cards). The '*' default hint gives partial routing help but no exclusions.

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