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

Annotations already carry the safety profile (destructiveHint=true, readOnlyHint=false), and the description adds value beyond them by disclosing exactly what gets destroyed ('stored outputs (classify group)'), irreversibility ('Cannot be undone'), and an unusual operational instruction to follow llmContext guidance in results. No contradiction with annotations — idempotentHint=true is consistent with an irreversible delete, and openWorldHint=true aligns with the cascading deletion scope.

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?

Three short sentences with zero filler: the core action is front-loaded, followed by the irreversibility warning, then the llmContext instruction. Each sentence carries distinct information and earns its place.

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?

With an output schema covering return values, rich parameter documentation, and annotations covering the safety profile, the description only needs to add behavioral facts — which it does (cascade scope, irreversibility, llmContext note). The one notable omission is the destructive-vs-cancel decision context, but an agent otherwise has what it needs to invoke the tool correctly.

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 100%, and the schema descriptions are rich: runId cites classify_document as provenance, environment explains the get_me API-key environment pinning, and workspaceId states the ws_... format plus the grant requirement. The tool description itself adds nothing parameter-specific, so the baseline 3 for high schema coverage applies.

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 states a specific verb+resource: 'Permanently delete a classify run and its stored outputs (classify group).' This clearly differentiates it from sibling delete_*_run tools (delete_extract_run, delete_split_run, delete_parse_run) by naming the exact resource type. The addition of the cascading 'classify group' scope makes the action precise with no ambiguity.

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?

No guidance is given on when to use this versus cancel_classify_run, which is the natural alternative for stopping or discarding a run without permanently destroying it and its outputs. The 'Cannot be undone' warning conveys stakes but not selection criteria, and the llmContext note concerns post-call behavior rather than tool choice. Given the sibling set contains both cancel_*_run and delete_*_run variants, the missing differentiation is a substantive gap.

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