Merge units

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

The annotations already declare readOnlyHint=false and destructiveHint=true, which covers the basic mutability and destructive nature. The description adds crucial behavioral context: it explains the confirmation token mechanism (call once to receive a token, then again with that token) and that the source unit is deleted. This goes beyond the annotations by detailing the workflow and the destructive effect on the source unit. However, it does not mention what happens to ingredient references if they cannot be repointed or if there are dependencies, but given the annotation coverage, the added detail is sufficient.

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 description is two sentences, front-loaded with the core action and then the critical confirmation requirement. Every word adds value; there is no filler or redundancy. It is concise yet complete for the key workflow.

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?

Given the tool's complexity (a destructive merge with a two-step confirmation), the description covers the essential workflow and parameter semantics. The schema is detailed, and annotations cover the destructive hint. However, it lacks information about error handling (e.g., what happens if the token is invalid, or if the units are already merged) and edge cases like merging a unit into itself. The absence of an output schema means the description could specify what the response contains (e.g., the token on first call), but it does not. Still, it is reasonably complete for an experienced user.

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?

The input schema already provides detailed descriptions for each parameter, including the UUID patterns and the meaning of 'from_id' (deleted) and 'to_id' (kept). The description reinforces this by explaining the overall repointing and deletion process, which adds semantic clarity to the parameters. The confirm_token parameter is also clearly explained in both schema and description. With 100% schema description coverage and this additional context, the description exceeds the baseline.

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 clearly states the function: 'Points every ingredient that uses one unit at another one and deletes the source unit.' This is a specific verb (merge) with a clear resource (units) and explicitly describes the outcome. It distinguishes itself from the sibling 'merge_foods' by focusing on units, and the detail about deleting the source unit clarifies its scope.

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 description provides context on how to use the tool: it requires a two-step confirmation process, which is a key usage guideline. However, it does not explicitly mention when to prefer this over alternatives like 'create_unit' or 'merge_foods', nor does it state any prerequisites (e.g., checking for existing ingredient references). The confirmation step is clear, but there is no guidance on what happens if the merge is invalid or if the units are in use.

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