undo_apply

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

Annotations already declare readOnlyHint=false, destructiveHint=true and idempotentHint=false; the description adds real value on top by explaining that the actual risk tier lives with the dispatched inverse tool, that dry_run previews without executing, and that the token is single-use. It does not describe what the inverse dispatch returns or what failure looks like, so it is not maximally transparent.

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 [WRITE][risk=medium] tag and the core behavior/constraint sentences are front-loaded and efficient, and the Args block is compact. The Args list partially restates schema parameter names, but each entry adds meaning, so waste is minimal.

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 non-idempotent, destructive, three-parameter mutation with no output schema, the description supplies the missing context an agent needs: the risk marker, the once-only token rule, the preview path, and parameter sourcing. Only the result shape of the dispatched inverse call is left unspecified, which is a minor 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 description coverage is 0%, so the description carries the full burden and largely meets it: undo_id is traced to its source (undo_list or `_undo_id` in a write result), dry_run is defined as a preview flag, and target is explained as pass-through to the inverse tool. The pass-through semantics of `target` (when it is rejected or ignored) remain somewhat vague.

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 gives a specific verb+resource ('Apply a recorded undo') and explains the mechanism (dispatching its inverse tool), which separates it cleanly from the sibling undo_list that only lists tokens. An agent can immediately tell what this does that its siblings do not.

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 states the context of use (apply a token obtained from undo_list or an `_undo_id` in a write result) and offers dry_run=True as the preview path, plus the constraint that a token can be applied only once. There is no explicit when-not-to-use or named alternative beyond undo_list, so it stops just short of a 5.

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