Tessl • The management layer for AI agent skills

A well-crafted router skill: dense keyword-to-skill routing tables, copy-paste commands, and workflows with explicit fallback and validation paths at every branch. The main improvements available are trimming the overloaded spawn-version block and moving long per-topic operational detail (version update procedure, agent-availability rules) into one-level-deep reference files.

Dimension	Reasoning	Score
Conciseness	The body is dense, table-driven, and assumes competence — no space is spent explaining concepts Claude already knows. However, the 「Spawn 版本提示」 blockquote is heavily overloaded (version-mismatch branches, notice strings, both a plain-language and technical-note instruction in one paragraph) and deploy/state reminders ("新开一个会话") repeat three times, so it fits 'efficient; minor instances that could be trimmed' rather than the lean level 5.	4 / 5
Actionability	Guidance is fully executable throughout: `node "<story-skill-dir>/scripts/dashboard-server.mjs" --root "<workspace>" --open`, `gh release view --json tagName,name,url -R zenstory-ai/oh-story-claudecode` with a curl fallback, exact three-line prompt templates for story-explorer/story-researcher, and per-host agent file paths (`.claude/agents/.md`, `.codex/agents/.toml`) — copy-paste ready commands covering the common cases.	5 / 5
Workflow Clarity	Every multi-step flow is explicitly sequenced with validation and error-recovery loops: check `node` before launching and wait for 「本机地址」; agent spawn preconditions with fallback to direct Read/Grep search; `gh` → `curl` → "查不到 → 告知…不报错"; the update flow is gated on user consent ("先报告版本，是否安装由用户决定") and dashboard saves verify content versions ("保存或确认删除时核对文件内容版本"). This matches the top anchor of explicit validation steps plus feedback loops.	5 / 5
Progressive Disclosure	Scored against the actual bundle: `scripts/dashboard-server.mjs` is referenced relatively and exists, and the dashboard's internals (assets/app.js, index.html, etc.) are correctly kept out of the overview; sections are clearly headed with no nested references. Minor gaps: long operational detail (the spawn/version-compat rules and the full 5-step update procedure) is inlined in the router file rather than split into a reference, and the referenced sibling `VERSION` file is absent from the bundle — good structure with minor organization gaps rather than the fully clean split.	4 / 5
	Total	18 / 20 Passed

Description

78%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 router-style description: it clearly states what it does and when to use it, with an explicit trigger list including natural Chinese user phrases. Its main weakness is that the capability surface is only sketched at the routing level — the concrete skills being routed to are never named — and a couple of triggers are generic enough to overlap other tooling.

Suggestions

Add one sentence enumerating the main routed capabilities (长篇/短篇写作、拆文、扫榜、去AI味、审稿、封面、导入) so the 'what' is comprehensive rather than just 'routes to skills'.

Include more natural trigger variants users would actually say, such as 「写网文」「拆书」「扫榜」「帮我审稿」「把我的书导进来」, to broaden keyword coverage.

Disambiguate the generic triggers (「打开工作台」「检查更新」) by tying them to this toolbox, e.g. 「打开网文工作台」「检查工具箱更新」, to reduce overlap risk with other dashboard/update tooling.

Dimension	Reasoning	Score
Specificity	"根据用户需求自动路由到对应 skill，并可启动本地 Dashboard" names the web-novel-toolbox domain with exactly two concrete actions (routing and launching the dashboard), but never enumerates the routed capabilities (拆文、扫榜、去味、审稿、封面 etc.), matching the '1-2 concrete actions, not comprehensive' anchor rather than the 'several specific actions' anchor.	3 / 5
Completeness	It explicitly answers both 'what' ("自动路由到对应 skill，并可启动本地 Dashboard") and 'when' with a dedicated 触发方式 clause listing concrete trigger phrases, matching the top anchor of clear what + explicit when with concrete triggers rather than the merely-explicit level 4.	5 / 5
Trigger Term Quality	Natural user phrases like 「我想写小说」「记住我的写作习惯」「打开工作台」「检查更新」 plus invocation forms /story、$story、/网文 give good coverage, but common variants users would say (e.g. 写网文、拆书、扫榜、审稿、导入旧书) are missing — good coverage with a few natural terms absent.	4 / 5
Distinctiveness Conflict Risk	The 网络小说 niche with triggers like 「我想写小说」 is mostly distinct with minimal conflict risk, but 「打开工作台」「检查更新」 are generic phrases that could match other dashboard/update tooling, fitting 'mostly distinct; minor overlap risk' rather than the fully distinct level 5.	4 / 5
	Total	16 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	14 / 16 Passed

Repository: zenstory-ai/oh-story-claudecode
Path: skills/story/SKILL.md
Revision: c3fbbbe

Reviewed: 4 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.

story

Table of Contents

Is this your skill?