novel_fix_plan

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

With no annotations, the description carries the full burden, and it does disclose two meaningful traits: it never produces prose (only direction), and audit reports only violations that can be evidenced, with fabricated line numbers/text rejected during audit_verify. What it omits is what the mark action actually mutates or persists, whether the plan is saved between calls, and whether any state file is written — significant gaps for a tool whose mark mode changes state.

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?

It is front-loaded with the core purpose and packed into a small footprint, so nothing is buried. But a large fraction of the text is a slash-delimited replay of the six action modes that reproduces the enum descriptions in the schema almost verbatim, and the run-on punctuation makes the mode list harder to scan than a structured list would be.

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 7-parameter, six-mode tool with no output schema, the description usefully sketches the return shape for plan items. It does not describe what verify, mark, audit, audit_brief, or audit_verify return, nor the preconditions for those modes, leaving the agent to guess at a substantial part of the contract.

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 100%, so the schema already defines every parameter, including the enums and the claims object shape, which establishes a baseline of 3. The description mostly restates the action enum and the itemId/state requirement rather than adding format, constraints, or edge-case meaning beyond the schema.

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 opening clause states a specific verb and resource: it converts a style diagnosis into a prioritized todo list, and it spells out the item shape (line range + original text, current vs target value, anchor paragraph, rewrite direction). It is clearly differentiable from generation-style siblings via the closing guarantee that it only gives direction, never sentences. It stops short of naming the diagnostic siblings (novel_diagnose / novel_style_check) it depends on, so the boundary is implied rather than stated.

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 enumerates all six action modes with their purposes, which does give real intra-tool routing guidance (plan → verify → mark as a workflow, audit_brief → audit_verify as a two-party loop). However, that mode rundown largely duplicates the action enum descriptions already present in the schema, and there is no guidance on when to use this tool instead of the sibling style/diagnosis tools, nor any stated prerequisite that a diagnosis must exist first.

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