Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/starfie1d1272/dsh-github-skills/githubnpx skills add Starfie1d1272/dsh-github-skills --skill githubgit clone --depth 1 https://github.com/Starfie1d1272/dsh-github-skillsWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00040 | $0.00530 |
| Opus 5 | $0.00020 | $0.00265 |
| Sonnet 5 | $0.00008 | $0.00106 |
| Haiku 4.5 | $0.00004 | $0.00053 |
Grade A, and why
github scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GitHub
General entrypoint for GitHub triage. Resolve context, classify the request, and route specialist work early. Do not re-implement specialist workflows here.
Capability policy
Use the most specific visible capability whose documented semantics cover
the need. Provider names and tool-name prefixes do not imply capability.
Use local git for local checkout facts; use gh only where no suitable
structured capability covers the operation. Never assume an unavailable
capability exists.
Routing
- general repository / issue / PR triage → stay here
- review feedback →
gh-address-comments - failing GitHub Actions →
gh-fix-ci - push / publish / open a PR — including fork contributions — with any
required branch/commit work →
gh-publish
When a specialist matches, load it with skill() and let it own the
workflow.
Mixed requests may require multiple specialists. Complete review or CI
domain work before publishing; gh-publish does not replace those
workflows.
Triage
- Resolve context. Repo, PR/issue number, or URL given by the user wins; for "this branch" / "the current PR", resolve the local git context first. Ask if still ambiguous.
- Gather the state relevant to the request (issue/PR metadata,
comments, checks) from visible capabilities or
gh. - Classify and handle here, or load the matching specialist and let it own the workflow.
- Report the state and the next action.
Scoped metadata actions (labels, reactions, issue edits) may be applied here only when explicitly requested; everything else routes.
Boundaries
- Flat comments are not review-thread state (see
gh-address-comments). - No CI-log claim without actual logs (see
gh-fix-ci). - Remote writes need explicit user intent or the host approval boundary.
Examples
- "Summarize the open PRs in this repo." → stay here
- "Address the review comments on PR 482." →
gh-address-comments - "Why are the GitHub Actions checks failing on this PR?" →
gh-fix-ci - "Fork awesome-foo, update its README, and open a PR." →
gh-publish
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 61 lines · 0 tokens per session scan A 0504fe9f1fbe
github is a skill published in the GitHub repository Starfie1d1272/dsh-github-skills (6 stars, last pushed 14d ago), licensed Apache-2.0. It adds 40 tokens to every session and 530 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
playlet-wechat-feed
短剧-公众号信息源 — 每日扫描公众号短剧爆款文章,按阅读量筛选热门内容,智能聚类题材方向后生成包含封面图、互动数据与创作洞察的HTML日报。支持按题材(穿越/霸总/重生等)、公众号、时间范围定向查询。⚠️查询前脚本先做输入校验：关键词需命中短剧题材词库（topickeywords…
playlet-bili-feed
B站短剧爆款内容追踪工具，每日自动扫描B站短剧创作内容，按点赞量筛选爆款作品，智能聚类题材后生成可视化HTML日报与创作趋势分析。.
playlet-xhs-feed
短剧-小红书信息源 — 每日扫描小红书短剧爆款内容,按互动量筛选热门笔记,智能聚类题材方向后生成包含封面、互动数据与创作洞察的HTML日报。支持按题材(穿越/霸总/重生等)、达人、时间范围定向查询。⚠️数据每日15:00更新前一天数据,目标日期无数据时必须先告知用户并等待确认后才能调用接口,禁止自动获取。⚠️查询前先校验分类/关键词是否符合短剧题材词库,不满足时不请求接口直接提示并推荐相关词。当用户需要短剧小红书日报、小红书短剧爆款、短剧热点、短剧创作趋势或自定义题材查询时使用。.
douyin-prohibited-word
检测文案、文件或网页中的抖音违禁词并加粗显示，提供违禁词替换建议和仅替换违禁词后的文案；当用户需要查询抖音平台违禁词、检查抖音笔记文案是否包含敏感词、或希望获得安全替换词和修改后文案时使用。.
douyin-subscribe
抖音账号订阅追踪 — 通过抖音号订阅账号（最多20个），Agent 每日 9:00 自动拉取并生成 HTML 报告。账号 ID 直接内置于自动化命令中，无需文件存储。支持多抖音号批量订阅，自动生成精美 HTML 作品报告，终端/Markdown 表格按账号依次展示作品标题、收藏、评论、分享、点赞、发布时间等数据。当用户订阅抖音账号、追踪抖音作品更新、监控抖音竞品账号时使用。触发词：抖音订阅、抖音账号订阅、抖音订阅追踪、抖音作品订阅、抖音账号监控、抖音作品追踪、抖音每日推送、抖音日报。.
global-ai-news-brief
全球AI新闻简报 — 一个关键词同时搜索抖音、小红书、公众号、B站、快手、视频号、今日头条、TikTok、Instagram、X(Twitter)、YouTube 共 11 大平台，跨平台聚合后由 AI 生成智能摘要、热点聚类、舆情分析和深度解读报告，输出终端表格 + 交互式 HTML 报告。包含详细信息源列表，适用于各种AI agent进行社媒资讯分析。当用户需要新闻搜索、舆情分析、热点追踪、跨平台情报对比、资讯聚合、新闻解读、全网搜索、事件追踪、行业情报时使用。触发词：新闻情报、全网搜索、舆情分析、热点追踪、新闻解读、资讯聚合、跨平台分析、事件追踪、行业情报、news brief、global news。.