CF clawhub-daily
每日扫描 ClawHub 全球 Skill 平台（500 个 Skill），通过 6 维度全维度推荐算法 为用户推荐 8 个有价值、不重复、值得关注的 AI Agent Skill，并通过多渠道推送完整简报。 数据出口说明（用户知情同意）： - 飞书（Lark）：云文档 + 卡片消息（有凭证时自动推送，无凭证时跳过） - 腾讯 IMA 知识库：官方 OpenAPI 推送（方式 A，有凭证时自动推送）；若安装了 ima-skill CLI 则自动调用（方式 B） - Obsidian 本地 vault：inbox/clawhub-daily/ 子目录（默认开启，仅写入本地磁盘）。vault 不可写时 fallback 到脚本目录 saved/ 子目录 - 本地文件：data/recommended/*.md 简报文件（默认开启，仅写入本地磁盘） 推送行为：executor 会尝试所有渠道，每个渠道独立 try/except，有凭证则推送，无凭证则跳过，一处失败不阻断其他渠道。 凭证来源（优先级：CLI 参数 > 环境变量 > config.json）： - 飞书：FEISHU_APP_ID / FEISHU_APP_SECRET / FEISHU_USER_OPEN_ID 环境变量，或 references/config.json - IMA：IMA_OPENAPI_CLIENTID / IMA_OPENAPI_APIKEY 环境变量，或 references/config.json - Obsidian：OBSIDIAN_VAULT_PATH 环境变量（默认 E:\Obsidian\md\inbox\clawhub-daily） 未配置任何凭证时仅生成本地文件，不执行外部推送。 本技能运行时不读取 GH_TOKEN、GITHUB_TOKEN 或其他与推荐功能无关的环境变量（GH_TOKEN 仅用于维护者发布新版本，见 docs/PUBLISHING_GUIDE.md）。 🔐 权限声明（capabilities）： - network（必需）：访问 ClawHub API（wry-manatee-359.convex.cloud）抓取数据；有凭证时访问飞书 API（open.feishu.cn）和 IMA API（ima.qq.com）推送简报 - filesystem（必需）：写入 data/snapshots/（原始数据）、data/recommended/（简报）、Obsidian vault inbox/clawhub-daily/（简报副本）、saved/（vault 不可写时的 fallback） - env_vars（可选）：FEISHU_APP_ID / FEISHU_APP_SECRET / FEISHU_USER_OPEN_ID / IMA_OPENAPI_CLIENTID / IMA_OPENAPI_APIKEY / OBSIDIAN_VAULT_PATH。未配置时仅生成本地文件 - subprocess（可选）：auto 模式下若官方 API 失败，可能调用 ima-skill CLI 命令（需用户预装） 不申请的权限：shell 执行（除 ima-skill CLI 外）、系统信息收集、GitHub 凭证读取（GH_TOKEN）、任意文件系统访问 触发场景： - 用户希望每日/定时收到 ClawHub Skill 推荐简报 - 用户希望跟踪 AI Agent 生态的最新 Skill 趋势 - 用户希望按痛点场景（自动化办公/开发工具/内容创作/数据采集/AI 增强/中文支持/金融分析）匹配推荐 - 用户希望避免重复推荐，结合 7 天跨维度去重 核心能力： - 真实抓取 ClawHub Top 500 Skill（基于 Convex API，0 token 消耗） - 适配新 API 数据结构（stats.installs / categories / badges.verified / changelog） - 6 维度全维度推荐：trending / quality / newcomers / panorama / actively_maintained / verified - 降级策略：候选不足时自动放宽阈值（fallback_fn） - optional 维度：verified 候选为 0 时自动跳过，不占配额 - 7 天跨维度去重，避免重复推荐 - 痛点匹配去重展示：同一 Skill 只在第一个命中场景展示 - changelog 展示：有最近变更摘要的 Skill 展示"最近变更"字段 - 痛点加权：基于 7 大场景库个性化排序 - 多渠道推送：飞书云文档 + Top 5 卡片消
每日扫描 ClawHub 全球 Skill 平台（500 个 Skill），通过 6 维度全维度推荐算法 为用户推荐 8 个有价值、不重复、值得关注的 AI Agent Skill，并通过多渠道推送完整简报。 数据出口说明（用户知情同意）： - 飞书（Lark）：云文档 +…
As a process F 36/100 · Will not run — References files that are not bundled: references/config.json, references/config.local.json
How to improve
- Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
- Shorten the description to 1024 characters.
- The text references files that are not there: add them or drop the references.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 19. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- error
description-longdescription is 2065 chars, limit 1024 - warning
description-no-whendescription does not say WHEN to use the skill (no "use when") - warning
missing-refreference to a missing file: references/config.json - warning
missing-refreference to a missing file: references/config.local.json - note
description-budgetdescription takes 2065 of the ~15000-char shared budget for all skills - note
frontmatter-keyunknown frontmatter key "slug" - note
frontmatter-keyunknown frontmatter key "displayName" - note
frontmatter-keyunknown frontmatter key "summary"
Process rating: all ten parameters 36/100
- 0Tools and files. 2 referenced file(s) missing: references/config.json, references/config.local.json
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 40Consistency. Frontmatter name (clawhub-daily) differs from the folder (skill-daily)
- 50Failures and branches. 0 branches, has a failure section
- 100Steps. 64 steps
- 100Execution cost. Instruction body is 1354 tokens
- 100Running it twice. No mutating operations
- low 14 top-level sections: this looks like several domains in one skill
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 2064: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- -214 emoji in the instructions: noise for the model
- -31 of 6 scripts are never mentioned in SKILL.md
- +2Single-language instructions
- +4Structure: 23 headings
- +3Step-by-step instructions: 64 items
- +4Has examples (3 code blocks)
- +4Reference files are cited in the instructions (6 of 8)
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 29.