novel_lexicon

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

With no annotations, the description carries the full behavioral burden and largely delivers: it discloses the two-layer storage with override semantics, the lexiconFirst default-on injection as a hard list, and safety behavior around mutations (force=false refuses to write to a corrupted layer to avoid silently wiping existing entries; overwrite defaults to true). It marks audit as read-only ('只读'). It omits permission/auth context and return-shape detail for write actions, keeping it short of a 5.

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 purpose and storage model are front-loaded, then actions are enumerated, then fields, then the lexiconFirst/injection note. For a 23-parameter, nine-action tool the length is defensible. It is a dense wall of text without formatting breaks, and the field enumeration partly duplicates the schema, so it is efficient but not maximally tight.

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?

There is no output schema and no annotations, so the description must do heavy lifting for a 23-param tool; it covers the storage model, per-action semantics, defaults (lexiconFirst on, overwrite true), failure handling (force), and audit output meaning (coverage plus line numbers and original sentences). It is close to complete, though return formats for list/export/scan are only lightly sketched, which is a minor gap given the absence of an output schema.

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 all 23 parameters are already documented in the schema, and the description's '字段：…' listing (term, kind, scene, avoid, aliases, note, priority) largely restates that. It does add modest cross-parameter meaning by tying parameters to actions (update 含改名 newTerm、单条停用; delete terms 或 all=true 清空某层; import replace=true 整层覆盖). Per the rubric, 100% coverage sets a baseline of 3, and the added value is only marginal.

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: a '必用词表' registry that stores scene-specific proper nouns and habitual words so the AI prefers them over generic phrasing. It clearly explains the two-layer storage (book-level overriding global _global) and enumerates all nine actions, so the tool's function is unambiguous. It stops short of explicitly distinguishing itself from the sibling novel_keywords, so it is clear but not fully differentiated.

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?

Each action carries inline usage guidance (list=查看, add=批量登记, audit=成稿用词审计, scan=从本书正文提取候选词), and it gives an explicit workflow cue: '写完一章用 action=audit 复核.' It also states that lexiconFirst defaults to on and that the list is injected as a hard checklist when non-empty. It never names when NOT to use this tool or routes the agent to a specific sibling (e.g., novel_keywords), so it lacks exclusion guidance.

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