curate-research

把一批 Harness Engineering 调研候选（文章/论文/工具的 URL）走完「抓取→翻译→评审→收录→清理」流水线，整合进本仓库 works/ 与 references/articles.md，并保持 C1–C13 一致性检查全绿。当用户说"处理这批调研候选 / 收录这些链接 / 整理 translate / 把这几篇翻译收进来"时使用。这是仓库给自己用的策展 harness。

Quality

84%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

73%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 body is a tightly structured operational runbook with strong workflow sequencing, validation checkpoints, and a copy-paste review prompt. Its main weaknesses are mild repetition between ④ and the 相关 section, and deferral of fetch/translate mechanics to external skills.

Suggestions

Remove the re-statement of the 减熵分流 rules and C13 explanation from the '相关' section, since both are already covered in ④ and its checklist — link to them instead of repeating.

Tighten the 2026-07-27 dated anecdote to the actionable rule ('数图要抓原文 HTML 数，可用 curl <url> | grep -c '<figure' 交叉验证') without the narrative setup.

For stages ① and ②, add one-line invocations or link targets for baoyu-url-to-markdown and baoyu-translate so a reader can execute them without resolving another skill's config.

Dimension	Reasoning	Score
Conciseness	The body is dense and mostly lean, but the '相关' section re-explains ④'s 减熵分流 rules and the C13 lint that were already covered in the checklist, and the dated 2026-07-27 anecdote runs long — material that could be tightened.	3 / 5
Actionability	Provides concrete commands (bash scripts/check-consistency.sh, curl <url> \| grep -c '<figure', cp to works/), a full copy-paste review prompt template, and a detailed pre-ingest checklist, but stages ① and ② defer mechanics to external baoyu skills/configs not detailed here.	4 / 5
Workflow Clarity	Six numbered phases with a diagram, an explicit non-skippable human gate, and validation checkpoints with feedback loops (check-consistency.sh must be green, checklist before ⑤, re-validate on failure) — matching the top anchor for a batch/destructive operation.	5 / 5
Progressive Disclosure	Well-organized into clear sections (何时用, 路径, 6 阶段, 计数同步清单, 相关) with no nested references and no bundle files; the long 计数同步清单 and lint details are inline but defensible for a single-session runbook, leaving only minor organization gaps.	4 / 5
	Total	16 / 20 Passed

Description

95%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 is strong: it states a concrete pipeline and explicit integration targets, and supplies verbatim natural-language trigger phrases for when to use it. Its only weakness is that the pipeline stages are compact labels rather than elaborated concrete actions.

Dimension	Reasoning	Score
Specificity	Names a concrete pipeline (抓取→翻译→评审→收录→清理) with concrete destinations (works/, references/articles.md) and C1–C13 checks, but the stages are compact one-word labels rather than fully spelled-out actions, leaving minor gaps versus the comprehensive anchor.	4 / 5
Completeness	Clearly states what the skill does (run the candidate pipeline, integrate into works/ and articles.md, keep C1–C13 green) and explicitly when to use it with concrete trigger phrases, matching the top anchor.	5 / 5
Trigger Term Quality	Explicitly quotes four natural user phrases a user would actually say ('处理这批调研候选 / 收录这些链接 / 整理 translate / 把这几篇翻译收进来'), giving comprehensive synonym-level trigger coverage.	5 / 5
Distinctiveness Conflict Risk	Targets a narrow, repo-specific self-curation niche (this repo's works/ + articles.md + C1–C13), making it highly distinct with minimal risk of triggering for the wrong skill.	5 / 5
	Total	19 / 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: 6 missing	Warning

	Total	15 / 16 Passed

Repository: deusyu/harness-engineering
Path: .claude/skills/curate-research/SKILL.md
Commit: 858c0da

Reviewed: 12 days ago

Table of Contents

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.