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

Annotations already declare non-read-only, non-idempotent, open-world, non-destructive, so the safety profile is covered; the description then adds substantial context beyond that: api_key provenance and bearer alternative, an explicit 422 failure mode for renamed fields, who signs in custody mode, and the ~60-second review floor before release for instant-classified claims. It stops short of describing behaviour on relay rejection or what the response body contains.

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 most important fact (what the tool does) is buried in the second clause after a credential preamble, and the text is a chain of em-dash-joined fragments mixing auth, payload format, custody models, timing, and follow-up tools. The content is dense but not front-loaded, so an agent must read the whole block to extract the core action.

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 five-parameter open-world mutation with no output schema, it explains the two payload branches and the release timing well, but omits the required lease_id, never says how a lease is acquired, and does not characterise failure or retry behaviour after a rejected relay. Useful but with a material gap.

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 only 40%, so the description must carry weight: it does explain doc_id+version pairing (including the 422 trap), raw_tx as a signed relay payload, and api_key semantics in detail. However lease_id — the single required parameter — is never mentioned anywhere in the description, leaving a required field with zero semantic help.

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?

The description conveys the effect only indirectly, through mode-specific instructions ('record your delivery doc version', 'platform signs submit() and relays it', 'pass raw_tx to relay the signed call'). No sentence states plainly that this submits a delivery against a claimed lease, and it never distinguishes itself from slot_revise or slot_post, so the agent has to infer both the verb and the boundary.

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 gives solid routing between operating modes (worker vs platform-custody vs self-custody), which is real when-to-use guidance. But it offers no guidance on choosing this tool over the sibling slot_revise or slot_post, and never states the precondition that a lease must first exist via slot_claim.

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