CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

docs-generator

Creates task-oriented technical documentation with progressive disclosure. Use when writing READMEs, API docs, architecture docs, or markdown documentation. Also use this skill at the END of any completed reverse engineering, penetration testing, CTF, or security analysis task to generate a formal report in the user's project directory. Trigger keywords: 写报告, 写文档, 出报告, writeup, 技术文档, report, documentation.

Quality

76%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/docs-generator/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

53%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The security-report half of the body is well-built: specific templates, a clear flavor-selection table, concrete output rules, and a final self-check. However, the ACTION REQUIRED block references a nonexistent '工作流' section and missing parent-level files, instructions are repeated across sections, and the generic documentation-writing half restates knowledge Claude already has.

Suggestions

Define an actual ordered 工作流 section (gather evidence → select template → pick flavor → draft → self-check) or remove the '进入工作流第一步' directive, since the body currently tells the reader to enter a workflow that does not exist.

Remove or fix broken parent-relative references (../tool-index.md, ../ops/evidence-finding-path.md, ../scripts/case-init.ps1) and reconcile the contradiction between 'ACTION REQUIRED' step 3 (call bootstrap when tools are missing) and '按需自举' stating no bootstrap is needed.

Deduplicate the vendor-report-rules/flavor guidance (stated in 模板选择, 厂商报告结构, 质量要求, and 路由上下文) into a single section, and trim the generic Core Principles/Formatting Standards/Anti-Patterns content that restates standard technical-writing knowledge.

DimensionReasoningScore

Conciseness

The security-report workflow is dense and specific, but there is real tightening to do: 'MUST 读取 references/vendor-report-rules.md' and the flavor rules are restated three times (模板选择, 质量要求, 路由上下文), the template file list is duplicated in 路由上下文, and ~50 lines of Core Principles/Formatting Standards/Anti-Patterns teach writing conventions Claude already knows. Matches 'mostly efficient but includes some unnecessary explanation'; not 2 because the security half is accurate, dense table-driven content rather than padded prose.

3 / 5

Actionability

Concrete elements exist (filename format 'YYYY-MM-DD_[类型]-[目标简称]-report.md', template-selection and flavor tables, output/language rules), but the ACTION REQUIRED block directs reading '../tool-index.md' which does not exist, step 4 says '进入"工作流"第一步' while no 工作流 section exists in the body, and '../ops/evidence-finding-path.md' and '../scripts/case-init.ps1' are also missing. Key execution details are dangling, matching 'some concrete guidance but incomplete'; not 4 because broken and dangling references are more than minor gaps.

3 / 5

Workflow Clarity

A rough sequence is distributed across sections (触发时机 → 模板选择 → 输出规范 → 任务完成自检 as a final validation checklist), but the steps are never assembled into one ordered workflow, intermediate checkpoints are implicit, and the headline 'ACTION REQUIRED' step points to a nonexistent '工作流' section. Matches 'steps listed but validation gaps; sequence present but checkpoints missing or implicit'; not 4 because the core execution path the reader is told to enter does not exist.

3 / 5

Progressive Disclosure

The three real bundle files (references/security-report-templates.md, templates.md, vendor-report-rules.md) are each referenced with clear purpose via the template-selection table and explicit MUST-read directives, one level deep — good structure. Not 5 because navigation is polluted by broken parent-relative references (../tool-index.md, ../ops/evidence-finding-path.md, ../scripts/case-init.ps1) and flavor/quality detail is inlined in the body instead of living in the references.

4 / 5

Total

13

/

20

Passed

Description

87%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description that explicitly covers what the skill does, when to use it, and a comprehensive bilingual trigger keyword list. Its main weakness is the generic 'report'/'documentation' triggers and the document-type-level action list, which create overlap risk with general writing/documentation skills.

DimensionReasoningScore

Specificity

Lists several concrete actions — 'writing READMEs, API docs, architecture docs, or markdown documentation' and 'generate a formal report in the user's project directory' — matching the 'several specific actions; minor gaps' anchor. Not 5 because the actions stay at document-type granularity and never state what the output actually contains (structure, evidence chain, diagrams).

4 / 5

Completeness

Clearly answers 'what' ('Creates task-oriented technical documentation with progressive disclosure', 'generate a formal report') and 'when' with two explicit 'Use when...' clauses plus concrete trigger keywords, exactly matching the top anchor. Not 4 because the 'when' guidance is already explicit and specific rather than needing more detail.

5 / 5

Trigger Term Quality

Explicit keyword list '写报告, 写文档, 出报告, writeup, 技术文档, report, documentation' plus 'Use when' phrasing gives comprehensive natural-term coverage with synonyms in both Chinese and English, matching the comprehensive-synonyms anchor. Not 4 because nothing a user would plausibly say when needing this skill is missing from the trigger set.

5 / 5

Distinctiveness Conflict Risk

The dual niche (task-oriented tech docs + security/RE report generation) is somewhat specific, but bare trigger keywords 'report' and 'documentation' would fire on nearly any writing or documentation request, overlapping with generic writer and doc skills — matching 'somewhat specific but could still overlap'. Not 4 because the generic half of the triggers creates real, not minor, overlap risk; not 2 because the security-report half is a clearly defined niche with distinct triggers.

3 / 5

Total

17

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

zhaoxuya520/reverse-skill

Path

skills/docs-generator/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.