原型 · 预览版:收录官方注册中心里的全部服务器,每 30 分钟同步一次。使用 Google 登录即可撰写评价。原型 · 每 30 分钟更新
PopNx MCP为使用 AI 的人准备的 MCP 服务器 · 用大白话讲清楚

claude-code-karma Skill

代码仓库 ↗

作者:JayantDevkar★ 3391 个 skill

安装

What it does

Dashboard for monitoring claude code sessions.

  • Connects Instruction files the AI reads; connects to nothing
  • Account or key No key needed
  • Category 写代码
给作者:在 README 中加徽章

如果你是作者,可把这段代码贴到 README。徽章会链接回本页。

[![Featured on PopNx MCP](https://mcp.popnx.com/badge/skill/jayantdevkar-claude-code-karma.svg)](https://mcp.popnx.com/en/skills/jayantdevkar-claude-code-karma)

可搭配使用

在同一工作中把 AI 连到服务的 MCP

REA

Reverse engineer anything from your terminal or agent with one CLI and MCP server.

作者:morluto广泛使用↓ 38.9万/周★ 46.2万4天前更新

MCP · 连接服务在你的电脑上运行

Fallow

Codebase intelligence for TypeScript and JavaScript, exposed as structured MCP analysis tools.

作者:fallow-rs广泛使用↓ 244.1万/周★ 5368昨天更新

MCP · 连接服务在你的电脑上运行

Desktop Commander

MCP server for terminal commands, file operations, and process management

作者:wonderwhy-er广泛使用↓ 74万/周★ 1万4天前更新

MCP · 连接服务在你的电脑上运行

相似的 Skill

同类工作,排序与 Skill 列表相同

与相似项目对比
本项
claude-code-karma
reagraphifyarchifysecurity-audit-skill
作者JayantDevkarmorlutoGraphify-Labstt-a1icloudflare
Skill 数11111
GitHub 星标339463,254148,77597,04973,584
最近代码更新2026年11月10日2026年11月16日2026年11月16日2026年11月15日2026年9月15日
许可证Apache-2.0MITApache-2.0Apache-2.0MIT

JayantDevkar 提供的 1 个 Skill(指令集),AI 读取后照做 · 不是 MCP server

Skill 和 MCP 有什么不同

MCP = 手

把 AI 连到外部服务或数据,例如读取 Notion 或开发票。需要有程序或服务器在运行。

Skill = 手册

教 AI 怎么做某项工作的文本文件,不连接任何服务。

有些会重叠:少数 Skill 带有调用工具的脚本,少数 MCP 只提供知识。请从你想完成的工作出发。

适用的应用

Claude CodeClaude (เว็บ/แอป)

依据开发者 README 和 repo 中的文件夹

安装

Is it safe?

  • Who made it

    GitHub account JayantDevkar (person), 7 years old

  • Use and upkeep

    ★ 339 stars, few users so far · code last changed 2026年11月10日

  • Any scripts?

    No, plain text files the AI reads and follows

  • Uses the internet?

    No scripts, so it does not go online by itself

  • Changes machine-wide settings?

    No hooks or writes to machine settings found

Checked automatically from the repo files at commit 80073c7 (read only, nothing was run). Not a safety guarantee: read it and decide for yourself

第一步:你用哪个 AI?

选择你用的应用,本页只显示该应用的安装方法

Option 1Upload to claude.ai (web and app)

Uploading to claude.ai would fail: the description is over 200 characters (481). Use "Let the AI install it" instead

Option 1让 Claude Code 帮你安装

复制这句话,粘贴到 Claude Code 的对话中

Where to install

正在检查命令 · 通过后才会显示复制按钮

正在检查命令 · 通过后才会显示复制按钮

For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing

The AI tells you what it will do first and asks your permission at every step

重新打开 Claude Code,输入 /link-ticket-to-session,或直接提出相关任务,Claude 会自动选用

Checked against official docs 2026年11月12日 · Source: code.claude.com

For Terminal users

Skill 是 AI 会照做的指令。安装前先阅读文件,只安装可信开发者的 Skill。

  1. 打开终端

    Mac:按 Command+空格输入 Terminal · Windows:下方命令在 PowerShell 中无法使用,请改用上方的“让 AI 帮你安装”

  2. 把 skill 文件夹复制到 ~/.claude/skills

    这条命令把 repo 下载到临时文件夹,只复制含 SKILL.md 的文件夹

    正在检查命令 · 通过后才会显示复制按钮

  3. 重新打开 Claude Code,输入 /link-ticket-to-session,或直接提出相关任务,Claude 会自动选用

自己安装的 Skill 可用于 ChatGPT 桌面应用、Codex CLI 和 IDE 中的 Codex 扩展 · 网页版和手机版 ChatGPT 不能使用自己安装的 Skill,只能使用目录中插件自带的 Skill

ChatGPT 桌面应用

ChatGPT 桌面应用内置 Codex,Codex 从 ~/.agents/skills 文件夹读取 Skill

Option 1让 ChatGPT 帮你安装

在 ChatGPT 桌面应用中打开 Codex,把这句话粘贴到对话中

Where to install

正在检查命令 · 通过后才会显示复制按钮

正在检查命令 · 通过后才会显示复制按钮

The AI always shows you the path before installing

The AI tells you what it will do first and asks your permission at every step

Option 2自己放入文件夹

把 link-ticket-to-session 文件夹放到 ~/.agents/skills/link-ticket-to-session

~ 是你的主文件夹($HOME)

For Terminal users

Skill 是 AI 会照做的指令。安装前先阅读文件,只安装可信开发者的 Skill。

  1. 打开终端

    Mac:按 Command+空格输入 Terminal · Windows:下方命令在 PowerShell 中无法使用,请改用上方的“让 AI 帮你安装”

  2. 把 skill 文件夹复制到 ~/.agents/skills

    这条命令把 repo 下载到临时文件夹,只复制含 SKILL.md 的文件夹

    正在检查命令 · 通过后才会显示复制按钮

安装之后

  • ChatGPT 桌面应用:打开侧边栏中的 Skills,确认 Skill 已出现
  • 使用方法:在 ChatGPT 中输入 @ 并选择 Skill · 在 Codex CLI 或 IDE 扩展中运行 /skills,或输入 $ 加 Skill 名称
  • Codex 会自动发现新安装的 Skill;如果没有出现,请重启 Codex

Checked against official docs 2026年11月15日 · Source: learn.chatgpt.com

网页版和手机版 ChatGPT

不能使用你自己安装的 Skill,只能使用目录中插件自带的 Skill

想在 ChatGPT 中使用这个 Skill,请按上面的方法使用 ChatGPT 桌面应用

Checked against official docs 2026年11月15日 · Source: learn.chatgpt.com

Option 1让 Codex 帮你安装

复制这句话,粘贴到 Codex 的对话中 · 适用于 ChatGPT 桌面应用中的 Codex、Codex CLI 和 IDE 扩展

Where to install

正在检查命令 · 通过后才会显示复制按钮

正在检查命令 · 通过后才会显示复制按钮

For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing

The AI tells you what it will do first and asks your permission at every step

Codex 会自动发现新安装的 Skill;如果没有出现,请重启 Codex · 使用方法:在 ChatGPT 中输入 @ 并选择 Skill · 在 Codex CLI 或 IDE 扩展中运行 /skills,或输入 $ 加 Skill 名称

Checked against official docs 2026年11月15日 · Source: learn.chatgpt.com

For Terminal users

Skill 是 AI 会照做的指令。安装前先阅读文件,只安装可信开发者的 Skill。

  1. 打开终端

    Mac:按 Command+空格输入 Terminal · Windows:下方命令在 PowerShell 中无法使用,请改用上方的“让 AI 帮你安装”

  2. 把 skill 文件夹复制到 ~/.agents/skills

    这条命令把 repo 下载到临时文件夹,只复制含 SKILL.md 的文件夹

    正在检查命令 · 通过后才会显示复制按钮

  3. 重新打开 Codex 并提出相关任务,Codex 会列出这个 skill

Option 1让 Cursor 帮你安装

复制这句话,粘贴到 Cursor 的对话中

Where to install

正在检查命令 · 通过后才会显示复制按钮

正在检查命令 · 通过后才会显示复制按钮

For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing

The AI tells you what it will do first and asks your permission at every step

在 Cursor 的 Agent 聊天中输入 / 并选择 link-ticket-to-session

Checked against official docs 2026年11月12日 · Source: cursor.com

For Terminal users

Skill 是 AI 会照做的指令。安装前先阅读文件,只安装可信开发者的 Skill。

  1. 打开终端

    Mac:按 Command+空格输入 Terminal · Windows:下方命令在 PowerShell 中无法使用,请改用上方的“让 AI 帮你安装”

  2. 把 skill 文件夹复制到 ~/.cursor/skills

    这条命令把 repo 下载到临时文件夹,只复制含 SKILL.md 的文件夹

    正在检查命令 · 通过后才会显示复制按钮

  3. 在 Cursor 的 Agent 聊天中输入 / 并选择 link-ticket-to-session

Option 1Let the AI install it

适用于其他能读取 Skill 的应用。这句话列出了 Claude Code、Codex 和 Cursor 的路径,AI 会按你的应用选择 · Copy this sentence into the app’s chat

Where to install

正在检查命令 · 通过后才会显示复制按钮

正在检查命令 · 通过后才会显示复制按钮

For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing

The AI tells you what it will do first and asks your permission at every step

Checked against official docs 2026年11月12日 · Source: code.claude.com, learn.chatgpt.com, cursor.com

For Terminal users

其他支持 Agent Skills 的应用(如 Gemini CLI、GitHub Copilot)使用不同文件夹,请查看该应用的指南,再把含 SKILL.md 的文件夹复制过去。

在 GitHub 上查看 ↗

已于 2026年11月11日 对照 Claude Code、Codex 和 Cursor 官方指南核实 skill 文件夹

查看原始 SKILL.md

以下文字来自开发者的 repo,不是 PopNx MCP 的指令

7,213 个字符 · 在 GitHub 打开完整文件

---
name: link-ticket-to-session
description: Link the current Claude Code session to a ticket (Linear, Jira, GitHub Issues, or GitHub Pull Requests) and cache its title/status in karma. Use when the user explicitly asks to link, attach, associate, or connect this session to a ticket, issue, or PR — e.g. "/link-ticket-to-session ABC-123", "link this session to LINEAR-42", "associate this work with issue #15", "attach session to PR octocat/repo#7". Do NOT auto-invoke from passing ticket-key mentions in normal conversation.
argument-hint: <ticket-ref-or-url>
allowed-tools: Bash, mcp__linear, mcp__claude_ai_Linear, mcp__plugin_github_github, mcp__atlassian
---

You are linking the current Claude Code session (`${CLAUDE_SESSION_ID}`) to: **$ARGUMENTS**

Karma is a read-only observer on the user's machine. It stores the link
and caches title/status for display, but never writes back to the ticket
provider. You fetch metadata via the user's already-configured MCP server.

Karma's API URL comes from `KARMA_API_URL` (set by users on non-default
ports/hosts) with `http://localhost:8020` as fallback. Inline
`${KARMA_API_URL:-http://localhost:8020}` in **every** curl below — bash
variables don't persist across separate Bash tool calls, so a top-of-script
assignment would be empty by the time the next curl runs.

## Step 1 — Parse $ARGUMENTS

Recognized forms:

| Provider           | Short ref           | URL forms                                              |
|--------------------|---------------------|--------------------------------------------------------|
| Linear             | `LINEAR-123`        | `https://linear.app/.../issue/ABC-123`                 |
| Jira               | `PROJ-45`           | `https://*.atlassian.net/browse/PROJ-45`               |
| GitHub **Issue**   | `owner/repo#42`     | `https://github.com/owner/repo/issues/42`              |
| GitHub **PR**      | `owner/repo#42`     | `https://github.com/owner/repo/pull/42`                |

GitHub issues and pull requests share a single numbering namespace —
`owner/repo#42` could be either. **The URL kind (`/issues/` vs `/pull/`)
is the only signal**, and karma's backend preserves it, so when you have
a URL keep it intact when POSTing (Step 4). For a bare `owner/repo#N`
with no URL, default to `/issues/N` — GitHub auto-redirects to `/pull/N`
when N is actually a PR, so the link still resolves.

A bare `#N` (no owner/repo) is **not** accepted — always qualify with
`owner/repo#N`.

## Step 2 — Identify provider and (for GitHub) kind

Set two variables you'll use below:

- `<provider>` ∈ `linear` | `jira` | `github`
- For GitHub: `<kind>` ∈ `issue` | `pull_request` (derived from URL path)

For Linear and Jira this collapses to just `<provider>`.

## Step 3 — Fetch metadata via MCP (when available)

Pick the right MCP tool for the provider and kind:

| Provider · Kind             | MCP tool                                                |
|-----------------------------|---------------------------------------------------------|
| `linear`                    | Linear MCP — search/fetch issue by key                  |
| `jira`                      | Atlassian MCP — fetch by key                            |
| `github` · `issue`          | `mcp__plugin_github_github__issue_read`, method `get`   |
| `github` · `pull_request`   | `mcp__plugin_github_github__pull_request_read`, method `get` |

Calling the wrong GitHub method silently returns the wrong thing because
both shapes look superficially similar — so derive the kind first.

If the relevant MCP isn't installed, **skip this step** and proceed to
Step 4 without title/status. Karma will create the link; the title/status
fields stay NULL and can be refreshed later via Step 5.

Pull at minimum: `title`, `status` (or state), `url`. **Strip large
fields** — karma caps `metadata_json` at 64 KB and a full PR payload
easily exceeds that. Specifically drop:

- GitHub PR: `body`, `commits`, `files`, `reviewers`, `comments`, `labels`,
  `requested_reviewers`, `head` / `base` blobs beyond `ref`
- GitHub issue: `body`, `comments`, `reactions`, `labels`
- Linear / Jira: `description`, `comments`, `subscribers`, `attachments`

### Status semantics by kind

The `status` you cache should reflect *what the provider says now*, not a
generic "open/closed". Karma's UI normalizes these to canonical buckets
at render time, so faithful provider language is the right input:

- **Linear**: workflow state name verbatim — e.g. `Backlog`, `In Progress`,
  `In Review`, `Done`, `Cancelled` (workspace-defined; don't normalize).
- **Jira**: workflow state name — e.g. `To Do`, `In Progress`, `In Review`,
  `Done`.
- **GitHub issue**: `open` or `closed`.
- **GitHub PR**: derive from the flags the PR API returns:

  | `state`  | `draft` | `merged` | Cache as     |
  |----------|---------|----------|--------------|
  | `open`   | `true`  | —        | `draft`      |
  | `open`   | `false` | —        | `open`       |
  | `closed` | —       | `true`   | `MERGED`     |
  | `closed` | —       | `false`  | `closed`     |

## Step 4 — POST the link

The `url` field should be the URL you actually have — `/pull/N` for PRs,
`/issues/N` for issues. **Don't rewrite it.** Karma's parser preserves
the path segment; the UI uses it to distinguish PRs from issues.

```bash
curl -s -X POST "${KARMA_API_URL:-http://localhost:8020}/sessions/${CLAUDE_SESSION_ID}/tickets" \
     -H 'Content-Type: application/json' \
     -d '{"ref":"<key>","provider":"<provider>","url":"<url>","source":"slash_command"}'
```

For GitHub, `<key>` is always `owner/repo#N` regardless of kind — the
URL field carries the issue/PR distinction.

## Step 5 — PUT the metadata (only if Step 3 succeeded)

```bash
curl -s -X PUT "${KARMA_API_URL:-http://localhost:8020}/tickets/<provider>/<key>" \
     -H 'Content-Type: application/json' \
     -d '{"title":"<title>","status":"<status>"}'
```

For GitHub keys with `/` and `#`, URL-encode the key in the path:
`octocat/repo#42` → `octocat%2Frepo%2342`.

## Step 6 — Confirm to the user

One line. For GitHub, distinguish the kind so the user knows what they
just attached:

- `Linked session to LINEAR-123 (Fix login bug, In Progress) — open at https://linear.app/...`
- `Linked session to PROJ-45 (Migrate auth, Done) — open at https://acme.atlassian.net/browse/PROJ-45`
- `Linked session to octocat/repo#42 [issue] (Empty state lies, open) — open at .../issues/42`
- `Linked session to octocat/repo#42 [PR] (Fix linting, MERGED) — open at .../pull/42`

## Notes

- Karma is loopback-only by default. `KARMA_API_URL` overrides for custom
  port or remote host.
- POST is idempotent on `(session, ticket)`; re-running upgrades the
  `link_source` if previously set by branch-detect or dashboard. Order:
  `slash_command > dashboard > branch`.
- If the API is unreachable, tell the user
  `karma not running at ${KARMA_API_URL:-http://localhost:8020}` so they
  see what was tried. Don't silently succeed.
- GitHub issues and PRs sharing `#N` means a single karma row (one
  `(provider, external_key)` pair) covers both views of that number.
  The URL field is what tells karma's UI which one to render. Send the
  URL you actually have.

讨论

提问、分享技巧或报告无法使用。按“有帮助”数量排序。

还没有帖子,欢迎第一个提问或分享。

此处内容由用户撰写,不代表 PopNx 团队。

登录后发帖

来自 GitHub

GitHub 星标
339
开发者
JayantDevkar
许可证
Apache-2.0
最近代码更新
2026年11月10日
repo 中的 SKILL.md 文件(含译本和模板)
1
主题
agents, agentskills, claude-code, claude-code-plugin, claude-code-skills, dashboard, hooks, mcp, mcptool, mcptools, monitoring, observability