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

With no annotations, the description must carry the behavioral burden. It discloses the output format (PNG) and the xref-based selection mechanism, but does not mention overwrite behavior, error cases, or file-access assumptions.

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 a single dense sentence that front-loads the core action and then adds only the necessary workflow context. Every part 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?

For a simple 3-parameter extraction tool, the description covers the workflow, source format, selection key, and output format. Minor gaps like overwrite behavior and invalid-xref handling are not critical for basic invocation.

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%, but the description compensates by mapping pdf_path to the source PDF, xref to the redact_scan-provided image reference, and output_path to the PNG output. This adds meaningful context beyond the bare parameter names.

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 names a specific operation—extract a specified embedded image from a PDF and save it as PNG—and clearly situates it within a workflow distinct from siblings like redact_scan and redact_apply.

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 explains when to use the tool: after redact_scan provides an image xref list, extract suspicious logos as templates. It does not explicitly list when-not-to-use cases or compare with alternatives, so it falls 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.

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

With no annotations, the description carries the behavioral disclosure burden. It usefully discloses that additions persist and automatically affect future scans, but it does not mention duplicate handling, reversibility, permissions, or error behavior.

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?

A single sentence with two clauses: the first states the core action and the second states the meaningful consequence. There is no filler or redundant restating of the tool name.

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 simple one-parameter tool with no output schema, the description is adequately complete: it says what to do, what counts as valid input, and the effect on future scans. Minor omissions like duplicate behavior are not critical for basic invocation.

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 schema provides no description for 'terms' (0% coverage), but the description's '确认过的敏感词' supplies the semantic meaning and the criterion for inclusion. For a single string-array parameter, this is sufficient, though limits or formatting constraints are not specified.

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 identifies the resource (glossary/dictionary) and the action (adding/persisting confirmed sensitive terms), distinguishing it from glossary_remove and glossary_list. The verb '沉淀' is idiomatic, but the meaning is clear enough for an agent to understand the tool's function.

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 appropriate input ('confirmed' sensitive words) and the consequence for future scans, providing implied usage context. However, it does not explicitly contrast with alternatives like glossary_remove, redact_apply, or note when not to use this tool.

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

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

No annotations are present, so the description carries the burden of behavioral disclosure. The verb '查看' clearly indicates a non-mutating read operation, and '持久化' plus 'scan 时自动并入候选' disclose the glossary's persistence and its relevance to scan behavior. This is sufficient for a simple list operation with an output schema.

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?

A single concise sentence communicates purpose, persistence, and connection to scan behavior. Every word earns its place, and the key action '查看' is front-loaded.

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 parameterless list tool with an output schema, the description is complete. It names the resource, states the read-only intent, and clarifies how the glossary relates to scans. No critical information is missing for correct selection and invocation.

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 has zero parameters, so the baseline is 4. The description correctly does not invent unnecessary parameter details; it focuses on what the tool returns access to. No parameter semantics are needed beyond this.

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 uses the specific verb '查看' (view) with a clear resource: the persistent sensitive-word glossary. It also distinguishes this from sibling tools like glossary_add and glossary_remove by indicating it is a read-only listing operation, and the parenthetical about scan behavior adds useful scope that goes beyond the tool name itself.

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 clear context: this glossary is the persistent dictionary that gets automatically merged into candidates during scan. This implies the tool should be used when an agent needs to inspect the active glossary that affects scanning. It does not explicitly enumerate when-not-to-use or mention alternatives, but the context is clear enough for a zero-parameter list tool.

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

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

没有注解提供行为信息，因此描述需完全承担披露责任。仅说“移除”表明这是一种变更操作，但未说明是否永久、是否可撤销、权限要求、对不存在词条的行为、批量删除的边界等关键副作用。

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?

描述只有一句话，没有冗余信息，动词和资源都放在最前面，结构上非常简洁。但这种简洁以牺牲行为信息为代价，因此不算完美。

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?

工具是删除类操作，且无注解、无输出 schema、无参数说明，描述却只有一句话，缺少操作前提、永久性说明、失败场景和返回值预期。对于一个执行破坏性操作的 1 参数工具，信息量明显不足。

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 描述覆盖率为 0%，且描述完全没有解释 terms 参数的语义、格式或匹配规则。虽然“词条”与 terms 存在字面关联，但未说明数组元素是否为精确匹配、为空时行为、重复词条如何处理等，描述未能弥补 schema 的空白。

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?

描述使用明确的动词“移除”和资源“词典词条”，表明这是从词典中删除条目的操作。与兄弟工具 glossary_add、glossary_list 在动词上形成自然区分，但未显式说明与它们的边界。

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?

没有说明何时使用此工具、何时应使用 glossary_add 或 glossary_list，也没有说明前置条件或替代方案。唯一能推断用途的只有动词“移除”和兄弟工具名称。

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

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

With no annotations, the description must carry behavioral disclosure; it does by warning that the operation is a true deletion and irreversible, and by conditioning invocation on confirmed terms. It could add more about whether the original file is overwritten or what happens on partial matches, but the destructive nature is clearly disclosed.

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 cover the core behavior, the required preceding workflow, and the one optional-parameter constraint. There is no filler or repetition of schema information.

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 destructive tool with no output schema and no annotations, the description supplies the essential safety context: irreversible deletion and the scan/confirmation prerequisite. It still omits return behavior and whether output_path must differ from input_path, but it is strong enough for an agent to call the tool safely in the intended workflow.

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 0%, so the description must compensate; it clarifies that terms are a confirmed list of sensitive words and that logo_templates only supports PDF. However, input_path and output_path are left entirely to their parameter names, with no format, relationship, or overwrite semantics explained.

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 opens with '执行脱敏（真删除，不可恢复）' — a specific verb and outcome: apply redaction, true deletion, irreversible. It also distinguishes itself from redact_scan by framing scan as a prerequisite rather than the action.

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 explicitly says terms must already be a confirmed sensitive-word list and that redact_scan should be called first, with user confirmation, before invoking this tool. It also states logo_templates is PDF-only, which tells the agent when the optional parameter is usable.

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

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

With no annotations, the description carries full behavioral disclosure. It is transparent about scope—checking multiple layers page-by-page and judging whether sensitive terms are fully removed—but it does not state whether the operation is read-only, whether it generates a report, or what side effects occur. For an audit tool this is a meaningful but incomplete profile.

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 entire description is one dense, front-loaded sentence. Every clause contributes: audit rigor, audit surfaces, judgment goal, and required timing. There is no filler or redundant restatement of the tool name.

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 should cover what the audit returns and what is required for invocation. It does not state the return value form (pass/fail, report, etc.), does not explain logo_templates, and does not disclose whether any file is modified. The purpose is clear, but the operational contract is incomplete.

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 must explain the parameters. It hints at 'terms' via '敏感词' and the file via '脱敏文件', but it never explicitly maps doc_path, terms, or logo_templates to their roles. 'logo_templates' is entirely unexplained.

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 '审计' (audit) applied to a well-defined resource—redacted documents—and lists exact audit surfaces ('对象流/文本层/元数据/修订记录/批注/OCR'). It also states the success criterion ('判定敏感词是否彻底移除'), which clearly separates this verification tool from sibling redaction/scan tools.

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 gives an explicit usage moment: run this audit before delivering any redacted file ('交付脱敏文件前必须通过本审计'). It does not explicitly name alternatives or say when not to use it, but the delivery gate is a clear contextual signal.

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

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

With no annotations, the description carries the full burden and does well: it discloses non-destructive behavior, automatic OCR on scanned PDFs, the output categories (candidate list, metadata leakage, format-related list), and the NER model prerequisite plus noise caveat. No contradictions with annotations exist.

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 compact sentences with the most important safety property (non-destructive) front-loaded, followed by formats, outputs, and the NER caveat. No filler.

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?

The description covers purpose, formats, outputs, and a key parameter caveat, which is substantial for a tool with no annotations or output schema. However, it omits explanations for four parameters and does not describe return structure details beyond the high-level list, leaving an agent to infer important call semantics.

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 must compensate for five parameters. It only adds meaning for enable_ner (model requirement, recommendation to disable). It does not explain keywords, ner_labels, enable_regex, or input_path beyond their property names.

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 (scan), a resource (documents), a key behavioral distinction (non-destructive, does not modify files), and the output (candidate sensitive items). This clearly differentiates it from siblings like redact_apply, which presumably applies changes.

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 clearly conveys the scanning context and supported document types, and gives an explicit conditional usage note (disable NER for standard/official documents due to noise). It does not explicitly name alternatives or say 'use redact_apply to apply redactions,' so it stops short of a full when/when-not comparison.

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