Tessl • The management layer for AI agent skills

An exemplary lean router: a dense step table, explicit validation gates and guardrails, and concrete executable commands. The only structural gap is that several existing reference files (schemas/, formulas/, patent_type_search.yaml) are not navigable from the body.

Dimension	Reasoning	Score
Conciseness	The ~58-line body is pure routing: a step table maps each step to its prompt file, background concepts (what a 交底书 or CNIPA is) are never explained, and even the search script's bundle- and repo-relative paths are given without connecting prose. Every token earns its place.	5 / 5
Actionability	Guidance is fully concrete: exact file to read per step ("Step 1 → prompts/intake.md"), named pre-write validators ("外观成文前 tools/check_design_views.py；结构案成文前 tools/check_source_parts.py"), and three copy-paste-ready bash commands (check_layout.py --decompose/--family/--matrix, check_scorecard.py) covering the fence validation cases.	5 / 5
Workflow Clarity	Steps 1–8 plus 5.5/迭代/旁路/交付 are explicitly sequenced in a table with named validation gates: the 三态门禁 before Step 6, Step 8 self-check before delivery, fence layout/scorecard scripts, and "未确认立项说明，不分件写多篇交底" gating the batch split, with correction_handler.md/iteration_context.md as error-recovery paths. The batch-operation validation cap does not apply because validation is present and explicit.	5 / 5
Progressive Disclosure	A clean hub-and-spoke layout with one-level-deep, clearly signaled references; the two references/ paths cited in the body (references/design_view_cnipa.md, references/scorecards/gbt42748_fence.yaml) exist in the bundle. However references/patent_type_search.yaml, references/formulas/, and references/schemas/ are never linked from the body, leaving part of the shipped bundle undiscoverable from SKILL.md.	4 / 5
	Total	19 / 20 Passed

Description

75%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.

The description states its capabilities concretely and comprehensively with a clearly distinct niche, but it omits any when-to-use trigger clause, capping completeness. Trigger keyword coverage is good, though a few common synonyms (技术交底书, 专利申请) would broaden natural discovery.

Suggestions

Append a trigger clause to the frontmatter description, e.g. 「用户提到交底书、专利挖掘、查新、实用新型、外观设计或 /patent-disclosure 时使用」, so the 'when' guidance lives in the description itself and completeness is no longer capped at 3.

Fold in common synonyms such as 技术交底书 and 专利申请/专利围栏 to widen natural trigger coverage toward the comprehensive anchor.

Dimension	Reasoning	Score
Specificity	"专利点挖掘、轻量查新与成文、保护型1+N专利布局" lists four concrete actions spanning the full workflow (mine, novelty search, draft, fence) across all three named patent types (发明/实用新型/外观设计). Coverage is comprehensive for the domain, matching the multiple-specific-actions anchor rather than the minor-gaps anchor below it.	5 / 5
Completeness	The "what" is explicit and multi-part (挖掘, 查新, 成文, 1+N 布局), but the description contains no "Use when…" clause or equivalent trigger guidance — triggers like "用户说专利挖掘、交底书、查新" appear only in the body's 何时用 section, not in the description. Per the judging guideline, missing explicit trigger guidance caps this dimension at 3.	3 / 5
Trigger Term Quality	"交底书", "查新", "实用新型", "外观设计", "专利布局" are exactly the terms a Chinese patent filer would say naturally. However common synonyms and variations such as "技术交底书", "专利申请", and "专利围栏" are absent, so keyword coverage is good but not comprehensive.	4 / 5
Distinctiveness Conflict Risk	"中国专利交底书" names an unmistakable niche, and its triggers (交底书, 查新, 实用新型, 外观设计) are domain-specific terms no neighboring skill would claim. Conflict risk with generic document or writing skills is minimal.	5 / 5
	Total	17 / 20 Passed

Validation

93%

Warnings & errors only

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

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
referenced_paths_exist	Referenced path issues: 1 deeper-than-1-level	Warning

	Total	15 / 16 Passed

Repository: handsomestWei/patent-disclosure-skill
Path: skills/patent-disclosure/SKILL.md
Revision: 6935a14

Reviewed: 27 days ago

Discovery Implementation Validation

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.

patent-disclosure

Table of Contents

Is this your skill?