CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

story-review

多视角对抗式审查。full/lean 模式在已部署 reviewer agents 时并行 spawn;缺失/异常 agents 或 spawn 失败时自动降级 solo,参考文件不可读时使用内置 rubric fallback。触发方式:/story-review、/审查、「审查一下」「帮我审一下」。

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

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

Quality

Content

88%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.

This is a well-engineered orchestration skill: the multi-phase workflow is explicitly sequenced with validation checkpoints and fallback loops, instructions are concrete and executable, and detail is appropriately delegated to one-level-deep reference files. The main cost is token weight — the inline fallback rubric pack and the dense single-sentence version-mismatch blockquote at the top make every invocation carry material that is redundant when the reference files are readable.

DimensionReasoningScore

Conciseness

The body is extremely dense with zero padding and no explanations of concepts Claude already knows, but the ~35-line inline 内置审查基准包 (embedded fallback rubric pack) duplicates content that exists in references/quality-rubric.md and rubrics/*.md and is loaded into context on every invocation even when those files are readable.

4 / 5

Actionability

The body provides copy-paste-ready commands (e.g. `node scripts/normalize-punctuation.js --check <正文文件...>`), exact canonical file paths, a fixed-format 技术备注 output line, a concrete YAML findings schema with S1–S4 severity definitions, and explicit platform-rubric selection tables — fully executable guidance for the common cases.

5 / 5

Workflow Clarity

Phases 0–4 are explicitly sequenced, with Phase 0 as a numbered preflight checklist that validates per-host agent definitions (frontmatter/TOML parsing, name matching) and defines explicit error-recovery loops (spawn failed → stop and re-run as solo; malformed/missing agents → solo), while all manuscript operations are report-only by design.

5 / 5

Progressive Disclosure

SKILL.md is an orchestration overview with a canonical-path table exposing nearly all reference and script files one level deep, and every referenced path was verified to exist in the bundle. Minor gaps: the embedded fallback rubric pack inlines content that belongs in a reference file, and the tracking path chains two levels deep (review-tracking.md → tracking-transaction.md → tracking-initialization.md).

4 / 5

Total

18

/

20

Passed

Description

66%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 has an explicit, well-formed trigger clause and accurately conveys the skill's orchestration behavior (parallel spawn, solo degradation, embedded fallback), which is above average. Its weaknesses are that it describes the plumbing rather than the product — neither the reviewed content (web novels) nor the concrete review capabilities (structure/character/prose/consistency findings with evidence) are stated — and its generic 审查 trigger phrases have overlap risk with other review skills.

Suggestions

State the subject matter and concrete review actions in the description, e.g. 「审查小说章节的结构、角色、文字与设定问题,给出附原文依据的修改建议」 — this would lift both specificity and completeness.

Disambiguate the triggers from generic review requests by adding story-specific phrases such as 「审查一下这章」「帮我审一下剧情」/「看这章有什么问题」.

Briefly mention the output (author-facing severity-ranked findings report) so users know what the skill returns, not just how it orchestrates.

DimensionReasoningScore

Specificity

"多视角对抗式审查" names the approach and the description lists several concrete mechanics ("full/lean 模式…并行 spawn", "自动降级 solo", "内置 rubric fallback"), but it never states the subject matter (novels/fiction) or the actual review actions (structure, character, prose, consistency findings) — a larger coverage gap than the 'minor gaps' of a 4.

3 / 5

Completeness

Both halves are present: an explicit trigger clause (触发方式 with concrete phrases) and a 'what' (多视角对抗式审查 with mode/fallback mechanics). The 'when' is fully explicit, but the 'what' is terse and omits what content gets reviewed and what output is produced, so it does not reach the fully concrete level of a 5.

4 / 5

Trigger Term Quality

"触发方式:/story-review、/审查、「审查一下」「帮我审一下」" provides natural phrases a Chinese-speaking user would actually say plus a dedicated slash command; common variations like 「帮我看看这章」 or 「有什么问题」 are missing.

4 / 5

Distinctiveness Conflict Risk

"审查一下" and "帮我审一下" are generic review requests, and the description prose never mentions 小说/stories, so in a mixed environment these triggers could plausibly match code-review-type skills; only the /story-review command clearly disambiguates, leaving real overlap risk with similar review skills.

3 / 5

Total

14

/

20

Passed

Validation

81%

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

Validation — 13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

13

/

16

Passed

Repository

zenstory-ai/oh-story-claudecode

Path

skills/story-review/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.