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

Annotations already flag destructiveHint=true, and the description goes well beyond them: it lists the exact categories of model entities removed, which are restored to the original beams, that it re-reads the model afterwards, and that tidy additionally deletes thickness records and the 'B2P mesh' group that the plugin's own Undo leaves behind. Resumability and the dry-run default are behaviors no annotation conveys.

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?

All content is relevant, but it is delivered as one very long run-on paragraph where resumability, tidy behavior, default run selection and preview/apply mode are interleaved. Front-loading the undo purpose is good, but the behavioral clauses would scan far better as separate sentences or bullets.

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 covers what an agent needs: default target selection, preview-versus-apply, what a partial failure returns ('says what is left'), and how to resume. Nothing material about invoking this tool safely is missing.

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?

With 67% schema coverage, the description compensates for the undocumented `apply` parameter (previews unless true; runs as a background job on the website) and reinforces `runId`'s default. `tidy` is already fully documented in the schema, so the description adds little there, keeping this just under a 5.

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 ('Undoes') and a specific resource (a b2p_convert run from the ledger), and enumerates exactly what is reverted versus restored. An agent can distinguish this from b2p_convert or the other *_undo siblings without opening the 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?

Explicit call conditions: default targets the latest not-undone run, it previews unless apply:true, and repeated calls with the same runId resume after a partial failure or a stopped background job. The only minor omission is naming b2p_convert directly as the paired tool, but the runId semantics make that obvious.

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