Resolve a partial launch: retry failed children, activate the succeeded, or discard

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

Annotations only declare readOnlyHint=false and openWorldHint=true, so the safety profile is thin; the description adds materially by flagging WRITE semantics, the confirm:true gate, and the important fact that there is no server-side dry run — meaning the preview is a client-side request echo, not a validated simulation. What it omits is the blast radius of each op: 'discard' presumably destroys child launches, and no reversibility or permission information is given.

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?

Two tight sentences, front-loaded with the purpose and the action set, followed by the endpoint and the critical write/confirm caveat. No filler, and the most decision-relevant constraint is placed where it will be read.

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 destructive-capable, nested-body write tool with no output schema and openWorldHint=true, the description covers the confirm gate and the absence of a dry run but leaves out the effects and reversibility of each op, what the preview response contains, and whether retrying is bounded. Enough to call it safely-ish, not enough to call it confidently.

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 75% and the schema itself documents id, confirm, and idempotencyKey, so the description's confirm note largely restates the field description. Its real contribution is mapping the three enumerated op values to their effects, but that mapping is only lexical ('retry failed' vs. 'discard') and does not explain consequences such as whether discard deletes children or what state activate_succeeded leaves them in.

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 concrete verb ('Resolve') on a concrete resource ('a partial launch') and enumerates the three possible actions (retry failed children, activate the succeeded, discard), so the agent knows exactly what outcome space it is choosing from. It also pins the endpoint POST /v1/crosspost/{id}/resolve. It does not explicitly contrast itself with the sibling crosspost_launch / crosspost_preview, so sibling differentiation is inferred rather than stated.

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 trigger condition ('a partial launch') is implied clearly enough that an agent can infer this is post-launch recovery, not initial launching. However, there is no explicit when-not-to-use guidance, no pointer to crosspost_launch for fresh launches or crosspost_preview for inspection, and no guidance on which of the three ops to pick in a given failure state.

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