[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M79P5AYCCQESNXD4F62SY7WT":3,"skill-same:01M79P5AYCCQESNXD4F62SY7WT":70},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":13,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M79P5AYCCQESNXD4F62SY7WT","mcp-integration",null,"Configures and integrates MCP servers in Claude Code plugins. Covers server types (stdio, SSE, HTTP, WebSocket), .mcp.json and plugin.json setup, CLAUDE_PLUGIN_ROOT path handling, MCP tool naming, pre-allowing tools in commands, OAuth and token authentication, and integration testing with \u002Fmcp and claude --debug. Use when adding an MCP server to a plugin, integrating MCP into Claude Code, configuring .mcp.json, setting up Model Context Protocol, or connecting an external service via stdio, SSE, HTTP, or WebSocket.","AndreHahm\u002Fandres-cc-marketplace","AndreHahm","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F80748019?v=4","01M79N2VZJT9MAJ7K695C9VXQ4","01M79N2W06PFX8BSCQ592M8K8K",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmcp-integration\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: mcp-integration\ndescription: >-\n  Configures and integrates MCP servers in Claude Code plugins. Covers server\n  types (stdio, SSE, HTTP, WebSocket), .mcp.json and plugin.json setup,\n  CLAUDE_PLUGIN_ROOT path handling, MCP tool naming, pre-allowing tools in\n  commands, OAuth and token authentication, and integration testing with \u002Fmcp\n  and claude --debug. Use when adding an MCP server to a plugin, integrating\n  MCP into Claude Code, configuring .mcp.json, setting up Model Context Protocol,\n  or connecting an external service via stdio, SSE, HTTP, or WebSocket.\nallowed-tools: Read Write Edit Glob\n---\n\n# MCP Integration for Claude Code Plugins\n\nModel Context Protocol (MCP) enables Claude Code plugins to integrate with external\nservices and APIs by providing structured tool access. Use MCP integration to expose\nexternal service capabilities as tools within Claude Code.\n\n**Key capabilities:**\n- Connect to external services (databases, APIs, file systems)\n- Provide 10+ related tools from a single service\n- Handle OAuth and complex authentication flows\n- Bundle MCP servers with plugins for automatic setup\n\n## Quick Start\n\n1. **Choose server type** — stdio (local process) or SSE\u002FHTTP\u002FWS (hosted service)\n2. **Create `.mcp.json`** at plugin root with server configuration (use `$CLAUDE_PLUGIN_ROOT` for all paths)\n3. **Document required env vars** in plugin README\n4. **Test locally** — run `\u002Fmcp` to verify server appears, then test a tool call\n5. **Pre-allow MCP tools** in relevant command frontmatter using exact tool names from `\u002Fmcp`\n6. **Test error cases** — connection failure and auth failure before publishing\n\n## When to Use\n\n- Adding a new MCP server to a Claude Code plugin\n- Connecting to an external API, database, or hosted service\n- Exposing 10+ related tools from a single backend service\n- Implementing OAuth or token-based authentication flows\n\n## When NOT to Use\n\n- For simple bash scripts or file operations — use command hooks instead\n- When fewer than 3–4 tools are needed — inline `allowed-tools: Bash(cmd:*)` is simpler\n- For read-only reference data — use `@file` injection in commands instead\n\n---\n\n## MCP Server Configuration Methods\n\nPlugins can bundle MCP servers in two ways: plugin-root `.mcp.json` or inline in `plugin.json`. Both auto-start when the plugin is enabled — use `${CLAUDE_PLUGIN_ROOT}` for all bundled server commands and configuration file paths.\n\n### Method 1: Dedicated .mcp.json (Recommended)\n\nCreate `.mcp.json` at plugin root:\n\n```json\n{\n  \"database-tools\": {\n    \"command\": \"${CLAUDE_PLUGIN_ROOT}\u002Fservers\u002Fdb-server\",\n    \"args\": [\"--config\", \"${CLAUDE_PLUGIN_ROOT}\u002Fconfig.json\"],\n    \"env\": {\n      \"DB_URL\": \"${DB_URL}\"\n    }\n  }\n}\n```\n\nBest for multiple servers; clearer separation of concerns; easier to maintain.\n\n### Method 2: Inline in plugin.json\n\nAdd `mcpServers` field to plugin.json:\n\n```json\n{\n  \"name\": \"my-plugin\",\n  \"version\": \"1.0.0\",\n  \"mcpServers\": {\n    \"plugin-api\": {\n      \"command\": \"${CLAUDE_PLUGIN_ROOT}\u002Fservers\u002Fapi-server\",\n      \"args\": [\"--port\", \"8080\"]\n    }\n  }\n}\n```\n\nBest for simple single-server plugins where one config file is preferred.\n\n---\n\n## MCP Server Types\n\n| Type | Transport | Best For | Auth |\n|------|-----------|----------|------|\n| stdio | Process stdin\u002Fstdout | Local tools, custom servers | Env vars |\n| SSE | HTTP streaming | Hosted services, cloud APIs | OAuth |\n| HTTP | REST | API backends, token auth | Tokens |\n| ws | WebSocket | Real-time, streaming | Tokens |\n\n### stdio (Local Process)\n\nExecute local MCP servers as child processes. Claude Code spawns, manages, and\nterminates the process; communicates via stdin\u002Fstdout.\n\n```json\n{\n  \"filesystem\": {\n    \"command\": \"npx\",\n    \"args\": [\"-y\", \"@modelcontextprotocol\u002Fserver-filesystem\", \"\u002Fallowed\u002Fpath\"],\n    \"env\": { \"LOG_LEVEL\": \"debug\" }\n  }\n}\n```\n\n### SSE (Server-Sent Events)\n\nConnect to hosted MCP servers. OAuth flows handled automatically by Claude Code —\nuser authenticates in browser on first use.\n\n```json\n{\n  \"asana\": {\n    \"type\": \"sse\",\n    \"url\": \"https:\u002F\u002Fmcp.asana.com\u002Fsse\"\n  }\n}\n```\n\n### HTTP (REST API)\n\nConnect with token authentication via request headers.\n\n```json\n{\n  \"api-service\": {\n    \"type\": \"http\",\n    \"url\": \"https:\u002F\u002Fapi.example.com\u002Fmcp\",\n    \"headers\": {\n      \"Authorization\": \"Bearer ${API_TOKEN}\"\n    }\n  }\n}\n```\n\n### WebSocket (Real-time)\n\nFor persistent connections with real-time data streaming.\n\n```json\n{\n  \"realtime-service\": {\n    \"type\": \"ws\",\n    \"url\": \"wss:\u002F\u002Fmcp.example.com\u002Fws\",\n    \"headers\": {\n      \"Authorization\": \"Bearer ${TOKEN}\"\n    }\n  }\n}\n```\n\nFor per-type deep dives, config fields, and edge cases, see `references\u002Fserver-types.md`.\nFor OAuth, token, and environment variable auth patterns, see `references\u002Fauthentication.md`.\n\n---\n\n## Environment Variable Expansion\n\nAll MCP configurations support environment variable substitution:\n\n- **`${CLAUDE_PLUGIN_ROOT}`** — always use for plugin file references in JSON configs (portability across installations). In shell environments, `$CLAUDE_PLUGIN_ROOT` is equivalent.\n- **User env vars** — e.g., `\"API_KEY\": \"${MY_API_KEY}\"` — passed from user's shell environment\n\nDocument all required environment variables in the plugin README.\n\n---\n\n## MCP Tool Naming\n\nTools from MCP servers are automatically prefixed:\n\n**Format:** `mcp__plugin_\u003Cplugin-name>_\u003Cserver-name>__\u003Ctool-name>`\n\n**Example:**\n- Plugin: `asana`, Server: `asana`, Tool: `create_task`\n- **Full name:** `mcp__plugin_asana_asana__asana_create_task`\n\nRun `\u002Fmcp` after connecting to see exact tool names.\n\n**Use the full name everywhere** — permission rules, skill `allowed-tools`, and agent `tools` — not just command frontmatter.\n\n### Pre-allowing in Commands\n\n```markdown\n---\nallowed-tools: [\n  \"mcp__plugin_asana_asana__asana_create_task\",\n  \"mcp__plugin_asana_asana__asana_search_tasks\"\n]\n---\n```\n\n**Wildcard (use sparingly):** `\"mcp__plugin_asana_asana__*\"` — prefer explicit tool names for security.\n\nFor patterns using MCP tools in commands and agents, see `references\u002Ftool-usage.md`.\n\n---\n\n## Lifecycle Management\n\n1. Plugin loads → MCP configuration parsed\n2. Server process started (stdio) or connection established (SSE\u002FHTTP\u002FWS)\n3. Tools discovered and registered as `mcp__plugin_...__...`\n4. Tools available in commands and agents\n\n**Notes:**\n- Use `\u002Fmcp` to see all registered servers and their tools\n- Restart Claude Code after configuration changes\n- stdio servers terminate when Claude Code exits\n\n---\n\n## Integration Patterns\n\n### Pattern 1: Simple Tool Wrapper\n\nCommand with validation before MCP call:\n\n```markdown\n---\nallowed-tools: [\"mcp__plugin_name_server__create_item\"]\n---\nGather item details from user, validate inputs, then use mcp__plugin_name_server__create_item.\n```\n\n### Pattern 2: Autonomous Agent\n\nAgent using MCP tools in a multi-step workflow without user interaction:\n\n```markdown\n---\nname: db-insights\ndescription: Use when the user asks to \"generate database report\" or \"analyze query data\"\nmodel: inherit\n---\n\nQuery recent data via mcp__plugin_myplug_database__query using the filters\nprovided. Analyze the results for trends and anomalies. Generate a structured\ninsights report with findings and recommendations. Present a summary to the user.\n```\n\n### Pattern 3: Multi-Server Plugin\n\nIntegrate multiple hosted services in one plugin's `.mcp.json`:\n\n```json\n{\n  \"github\": { \"type\": \"sse\", \"url\": \"https:\u002F\u002Fmcp.github.com\u002Fsse\" },\n  \"jira\":   { \"type\": \"sse\", \"url\": \"https:\u002F\u002Fmcp.jira.com\u002Fsse\" }\n}\n```\n\nCommand that uses tools from both servers:\n\n```markdown\n---\nallowed-tools: [\n  \"mcp__plugin_devtools_github__create_pull_request\",\n  \"mcp__plugin_devtools_jira__create_issue\"\n]\n---\nCreate a GitHub PR for the staged changes, then open a linked Jira issue for\ntracking. Use mcp__plugin_devtools_github__create_pull_request with branch\ndetails, then mcp__plugin_devtools_jira__create_issue with the PR URL in the\ndescription field.\n```\n\n---\n\n## Security Best Practices\n\n**DO:**\n- Use `${CLAUDE_PLUGIN_ROOT}` for all plugin file path references\n- Use environment variables for tokens: `\"Authorization\": \"Bearer ${API_TOKEN}\"`\n- Use HTTPS\u002FWSS for all remote connections\n- Pre-allow specific MCP tools in commands (not wildcards)\n- Document all required env vars in README\n\n**DON'T:**\n- Hardcode tokens or credentials in configuration files\n- Commit credentials to git\n- Use `http:\u002F\u002F` or `ws:\u002F\u002F` for remote MCP servers\n- Use `mcp__plugin_...__*` wildcards unless truly necessary\n\n---\n\n## Troubleshooting\n\n| Symptom | Check |\n|---------|-------|\n| Server not connecting | URL correct? Server running (stdio)? Network accessible? |\n| Tools missing from `\u002Fmcp` | Server connected? Restart Claude Code after config change. |\n| Auth failing | Env vars set? Token scope correct? Clear cached OAuth tokens and retry. |\n| `${CLAUDE_PLUGIN_ROOT}` not expanding | `.mcp.json` at plugin root? Plugin installed (not just cloned)? |\n| Tool call errors | Validate inputs. Check rate limits \u002F quotas. Review `claude --debug` logs. |\n\nEnable debug logging with `claude --debug` — look for MCP server connection, tool discovery, and auth flow log entries.\n\n---\n\n## Testing & Validation\n\nAfter configuring MCP integration:\n\n1. **JSON syntax** — `jq empty .mcp.json && echo \"Valid\"` before testing anything else\n2. **Server registration** — run `\u002Fmcp`; confirm server name and all expected tools appear\n3. **Tool call test** — invoke a command that uses an MCP tool; verify it succeeds end-to-end\n4. **Auth test** — for OAuth: trigger first-use browser flow; for tokens: confirm env var is read\n5. **Error case** — stop the server (stdio) or use a bad URL; confirm a clear failure message\n\n**Quality gates:**\n- [ ] `\u002Fmcp` shows server and all expected tools\n- [ ] At least one tool call succeeds end-to-end\n- [ ] Required env vars documented in README\n- [ ] No hardcoded credentials in `.mcp.json` or `plugin.json`\n- [ ] HTTPS\u002FWSS used for all remote servers (not HTTP\u002FWS)\n- [ ] Specific tools pre-allowed in commands (not wildcards)\n\n---\n\n## Reference Guide\n\n| Resource | Purpose |\n|---|---|\n| `references\u002Fserver-types.md` | Deep dive on each server type — config fields, edge cases |\n| `references\u002Fauthentication.md` | OAuth, token, and env var auth patterns |\n| `references\u002Ftool-usage.md` | Using MCP tools in commands and agents |\n| `examples\u002Fstdio-server.json` | Working stdio server configuration |\n| `examples\u002Fsse-server.json` | Working SSE server with OAuth |\n| `examples\u002Fhttp-server.json` | Working HTTP server with token auth |\n","# MCP Integration for Claude Code Plugins\n\nModel Context Protocol (MCP) enables Claude Code plugins to integrate with external\nservices and APIs by providing structured tool access. Use MCP integration to expose\nexternal service capabilities as tools within Claude Code.\n\n**Key capabilities:**\n- Connect to external services (databases, APIs, file systems)\n- Provide 10+ related tools from a single service\n- Handle OAuth and complex authentication flows\n- Bundle MCP servers with plugins for automatic setup\n\n## Quick Start\n\n1. **Choose server type** — stdio (local process) or SSE\u002FHTTP\u002FWS (hosted service)\n2. **Create `.mcp.json`** at plugin root with server configuration (use `$CLAUDE_PLUGIN_ROOT` for all paths)\n3. **Document required env vars** in plugin README\n4. **Test locally** — run `\u002Fmcp` to verify server appears, then test a tool call\n5. **Pre-allow MCP tools** in relevant command frontmatter using exact tool names from `\u002Fmcp`\n6. **Test error cases** — connection failure and auth failure before publishing\n\n## When to Use\n\n- Adding a new MCP server to a Claude Code plugin\n- Connecting to an external API, database, or hosted service\n- Exposing 10+ related tools from a single backend service\n- Implementing OAuth or token-based authentication flows\n\n## When NOT to Use\n\n- For simple bash scripts or file operations — use command hooks instead\n- When fewer than 3–4 tools are needed — inline `allowed-tools: Bash(cmd:*)` is simpler\n- For read-only reference data — use `@file` injection in commands instead\n\n---\n\n## MCP Server Configuration Methods\n\nPlugins can bundle MCP servers in two ways: plugin-root `.mcp.json` or inline in `plugin.json`. Both auto-start when the plugin is enabled — use `${CLAUDE_PLUGIN_ROOT}` for all bundled server commands and configuration file paths.\n\n### Method 1: Dedicated .mcp.json (Recommended)\n\nCreate `.mcp.json` at plugin root:\n\n```json\n{\n  \"database-tools\": {\n    \"command\": \"${CLAUDE_PLUGIN_ROOT}\u002Fservers\u002Fdb-server\",\n    \"args\": [\"--config\", \"${CLAUDE_PLUGIN_ROOT}\u002Fconfig.json\"],\n    \"env\": {\n      \"DB_URL\": \"${DB_URL}\"\n    }\n  }\n}\n```\n\nBest for multiple servers; clearer separation of concerns; easier to maintain.\n\n### Method 2: Inline in plugin.json\n\nAdd `mcpServers` field to plugin.json:\n\n```json\n{\n  \"name\": \"my-plugin\",\n  \"version\": \"1.0.0\",\n  \"mcpServers\": {\n    \"plugin-api\": {\n      \"command\": \"${CLAUDE_PLUGIN_ROOT}\u002Fservers\u002Fapi-server\",\n      \"args\": [\"--port\", \"8080\"]\n    }\n  }\n}\n```\n\nBest for simple single-server plugins where one config file is preferred.\n\n---\n\n## MCP Server Types\n\n| Type | Transport | Best For | Auth |\n|------|-----------|----------|------|\n| stdio | Process stdin\u002Fstdout | Local tools, custom servers | Env vars |\n| SSE | HTTP streaming | Hosted services, cloud APIs | OAuth |\n| HTTP | REST | API backends, token auth | Tokens |\n| ws | WebSocket | Real-time, streaming | Tokens |\n\n### stdio (Local Process)\n\nExecute local MCP servers as child processes. Claude Code spawns, manages, and\nterminates the process; communicates via stdin\u002Fstdout.\n\n```json\n{\n  \"filesystem\": {\n    \"command\": \"npx\",\n    \"args\": [\"-y\", \"@modelcontextprotocol\u002Fserver-filesystem\", \"\u002Fallowed\u002Fpath\"],\n    \"env\": { \"LOG_LEVEL\": \"debug\" }\n  }\n}\n```\n\n### SSE (Server-Sent Events)\n\nConnect to hosted MCP servers. OAuth flows handled automatically by Claude Code —\nuser authenticates in browser on first use.\n\n```json\n{\n  \"asana\": {\n    \"type\": \"sse\",\n    \"url\": \"https:\u002F\u002Fmcp.asana.com\u002Fsse\"\n  }\n}\n```\n\n### HTTP (REST API)\n\nConnect with token authentication via request headers.\n\n```json\n{\n  \"api-service\": {\n    \"type\": \"http\",\n    \"url\": \"https:\u002F\u002Fapi.example.com\u002Fmcp\",\n    \"headers\": {\n      \"Authorization\": \"Bearer ${API_TOKEN}\"\n    }\n  }\n}\n```\n\n### WebSocket (Real-time)\n\nFor persistent connections with real-time data streaming.\n\n```json\n{\n  \"realtime-service\": {\n    \"type\": \"ws\",\n    \"url\": \"wss:\u002F\u002Fmcp.example.com\u002Fws\",\n    \"headers\": {\n      \"Authorization\": \"Bearer ${TOKEN}\"\n    }\n  }\n}\n```\n\nFor per-type deep dives, config fields, and edge cases, see `references\u002Fserver-types.md`.\nFor OAuth, token, and environment variable auth patterns, see `references\u002Fauthentication.md`.\n\n---\n\n## Environment Variable Expansion\n\nAll MCP configurations support environment variable substitution:\n\n- **`${CLAUDE_PLUGIN_ROOT}`** — always use for plugin file references in JSON configs (portability across installations). In shell environments, `$CLAUDE_PLUGIN_ROOT` is equivalent.\n- **User env vars** — e.g., `\"API_KEY\": \"${MY_API_KEY}\"` — passed from user's shell environment\n\nDocument all required environment variables in the plugin README.\n\n---\n\n## MCP Tool Naming\n\nTools from MCP servers are automatically prefixed:\n\n**Format:** `mcp__plugin_\u003Cplugin-name>_\u003Cserver-name>__\u003Ctool-name>`\n\n**Example:**\n- Plugin: `asana`, Server: `asana`, Tool: `create_task`\n- **Full name:** `mcp__plugin_asana_asana__asana_create_task`\n\nRun `\u002Fmcp` after connecting to see exact tool names.\n\n**Use the full name everywhere** — permission rules, skill `allowed-tools`, and agent `tools` — not just command frontmatter.\n\n### Pre-allowing in Commands\n\n```markdown\n---\nallowed-tools: [\n  \"mcp__plugin_asana_asana__asana_create_task\",\n  \"mcp__plugin_asana_asana__asana_search_tasks\"\n]\n---\n```\n\n**Wildcard (use sparingly):** `\"mcp__plugin_asana_asana__*\"` — prefer explicit tool names for security.\n\nFor patterns using MCP tools in commands and agents, see `references\u002Ftool-usage.md`.\n\n---\n\n## Lifecycle Management\n\n1. Plugin loads → MCP configuration parsed\n2. Server process started (stdio) or connection established (SSE\u002FHTTP\u002FWS)\n3. Tools discovered and registered as `mcp__plugin_...__...`\n4. Tools available in commands and agents\n\n**Notes:**\n- Use `\u002Fmcp` to see all registered servers and their tools\n- Restart Claude Code after configuration changes\n- stdio servers terminate when Claude Code exits\n\n---\n\n## Integration Patterns\n\n### Pattern 1: Simple Tool Wrapper\n\nCommand with validation before MCP call:\n\n```markdown\n---\nallowed-tools: [\"mcp__plugin_name_server__create_item\"]\n---\nGather item details from user, validate inputs, then use mcp__plugin_name_server__create_item.\n```\n\n### Pattern 2: Autonomous Agent\n\nAgent using MCP tools in a multi-step workflow without user interaction:\n\n```markdown\n---\nname: db-insights\ndescription: Use when the user asks to \"generate database report\" or \"analyze query data\"\nmodel: inherit\n---\n\nQuery recent data via mcp__plugin_myplug_database__query using the filters\nprovided. Analyze the results for trends and anomalies. Generate a structured\ninsights report with findings and recommendations. Present a summary to the user.\n```\n\n### Pattern 3: Multi-Server Plugin\n\nIntegrate multiple hosted services in one plugin's `.mcp.json`:\n\n```json\n{\n  \"github\": { \"type\": \"sse\", \"url\": \"https:\u002F\u002Fmcp.github.com\u002Fsse\" },\n  \"jira\":   { \"type\": \"sse\", \"url\": \"https:\u002F\u002Fmcp.jira.com\u002Fsse\" }\n}\n```\n\nCommand that uses tools from both servers:\n\n```markdown\n---\nallowed-tools: [\n  \"mcp__plugin_devtools_github__create_pull_request\",\n  \"mcp__plugin_devtools_jira__create_issue\"\n]\n---\nCreate a GitHub PR for the staged changes, then open a linked Jira issue for\ntracking. Use mcp__plugin_devtools_github__create_pull_request with branch\ndetails, then mcp__plugin_devtools_jira__create_issue with the PR URL in the\ndescription field.\n```\n\n---\n\n## Security Best Practices\n\n**DO:**\n- Use `${CLAUDE_PLUGIN_ROOT}` for all plugin file path references\n- Use environment variables for tokens: `\"Authorization\": \"Bearer ${API_TOKEN}\"`\n- Use HTTPS\u002FWSS for all remote connections\n- Pre-allow specific MCP tools in commands (not wildcards)\n- Document all required env vars in README\n\n**DON'T:**\n- Hardcode tokens or credentials in configuration files\n- Commit credentials to git\n- Use `http:\u002F\u002F` or `ws:\u002F\u002F` for remote MCP servers\n- Use `mcp__plugin_...__*` wildcards unless truly necessary\n\n---\n\n## Troubleshooting\n\n| Symptom | Check |\n|---------|-------|\n| Server not connecting | URL correct? Server running (stdio)? Network accessible? |\n| Tools missing from `\u002Fmcp` | Server connected? Restart Claude Code after config change. |\n| Auth failing | Env vars set? Token scope correct? Clear cached OAuth tokens and retry. |\n| `${CLAUDE_PLUGIN_ROOT}` not expanding | `.mcp.json` at plugin root? Plugin installed (not just cloned)? |\n| Tool call errors | Validate inputs. Check rate limits \u002F quotas. Review `claude --debug` logs. |\n\nEnable debug logging with `claude --debug` — look for MCP server connection, tool discovery, and auth flow log entries.\n\n---\n\n## Testing & Validation\n\nAfter configuring MCP integration:\n\n1. **JSON syntax** — `jq empty .mcp.json && echo \"Valid\"` before testing anything else\n2. **Server registration** — run `\u002Fmcp`; confirm server name and all expected tools appear\n3. **Tool call test** — invoke a command that uses an MCP tool; verify it succeeds end-to-end\n4. **Auth test** — for OAuth: trigger first-use browser flow; for tokens: confirm env var is read\n5. **Error case** — stop the server (stdio) or use a bad URL; confirm a clear failure message\n\n**Quality gates:**\n- [ ] `\u002Fmcp` shows server and all expected tools\n- [ ] At least one tool call succeeds end-to-end\n- [ ] Required env vars documented in README\n- [ ] No hardcoded credentials in `.mcp.json` or `plugin.json`\n- [ ] HTTPS\u002FWSS used for all remote servers (not HTTP\u002FWS)\n- [ ] Specific tools pre-allowed in commands (not wildcards)\n\n---\n\n## Reference Guide\n\n| Resource | Purpose |\n|---|---|\n| `references\u002Fserver-types.md` | Deep dive on each server type — config fields, edge cases |\n| `references\u002Fauthentication.md` | OAuth, token, and env var auth patterns |\n| `references\u002Ftool-usage.md` | Using MCP tools in commands and agents |\n| `examples\u002Fstdio-server.json` | Working stdio server configuration |\n| `examples\u002Fsse-server.json` | Working SSE server with OAuth |\n| `examples\u002Fhttp-server.json` | Working HTTP server with token auth |\n",{"name":5,"description":7,"allowed-tools":23},"Read Write Edit Glob","main","1bf12900cac6f28e346abd2912bdb78d306de723",".claude\u002Fskills\u002Fmcp-integration","https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace","https:\u002F\u002Fgithub.com\u002FAndreHahm","2026-11-15 04:26:16",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add AndreHahm\u002Fandres-cc-marketplace --skill mcp-integration",{"tool":36,"command":37},"bunx","bunx skills add AndreHahm\u002Fandres-cc-marketplace --skill mcp-integration",{"tool":39,"command":40},"pnpm","pnpm dlx skills add AndreHahm\u002Fandres-cc-marketplace --skill mcp-integration","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmcp-integration\u002FSKILL.md",true,[44,50,54,58,62,66],{"path":45,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":48,"sourceHtmlUrl":49},"examples\u002Fhttp-server.json","REFERENCE","参考文档",502,"https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmcp-integration\u002Fexamples\u002Fhttp-server.json",{"path":51,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":52,"sourceHtmlUrl":53},"examples\u002Fsse-server.json",413,"https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmcp-integration\u002Fexamples\u002Fsse-server.json",{"path":55,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":56,"sourceHtmlUrl":57},"examples\u002Fstdio-server.json",816,"https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmcp-integration\u002Fexamples\u002Fstdio-server.json",{"path":59,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":60,"sourceHtmlUrl":61},"references\u002Fauthentication.md",9677,"https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmcp-integration\u002Freferences\u002Fauthentication.md",{"path":63,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":64,"sourceHtmlUrl":65},"references\u002Fserver-types.md",10753,"https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmcp-integration\u002Freferences\u002Fserver-types.md",{"path":67,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":68,"sourceHtmlUrl":69},"references\u002Ftool-usage.md",12370,"https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fmcp-integration\u002Freferences\u002Ftool-usage.md",[71,78,85,92,99,106],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":18},"01M79R0K918Z0XDM74R0EY5H3P","demo","Fixture skill used by scripts.marketplace_ci sync tests.","https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002Ftests\u002Fmarketplace_ci\u002Ffixtures\u002Fplugins\u002Fsample-kit\u002Fskills\u002Fdemo\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":18},"01M79R08VMDJA2FPJKH3ZM309K","export-demo","Fixture skill used by scripts.marketplace_ci export tests.","https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002Ftests\u002Fmarketplace_ci\u002Ffixtures\u002Fclaude\u002Fskills\u002Fexport-demo\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":18},"01M79R042FS581DS2JNW5AZ1MK","work-to-development","Validate an accepted Linear Issue's readiness for implementation (criteria, dependencies, duplicate\u002Fexisting artifacts) and request governed branch\u002Fworktree creation through git-kit:starting-work — recording work-started Git\u002FGitHub evidence and a deliberate Linear Started transition only after Git identity is read back and verified. Use when asked to start work on a Linear issue, begin development for an accepted issue, or prepare an accepted issue for implementation. Never creates a branch directly — always delegates to git-kit.","https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002Fplugins\u002Fworkmanagement-kit\u002Fskills\u002Fwork-to-development\u002FSKILL.md",[],[],{"skillNo":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":98,"publishedAt":18},"01M79QZZD4EHB82P72P4N92BKT","work-linking","Create and maintain stable cross-system links and authority labels between Notion and Linear records, and detect\u002Frepair drift between them. Use when asked to link a Notion record to a Linear record, check for drift between Notion and Linear, or repair a broken\u002Fstale link. Repairs only the bounded non-authoritative copy or link — never chooses by newest timestamp, and never creates a reverse-write loop against either system's own authority.","https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002Fplugins\u002Fworkmanagement-kit\u002Fskills\u002Fwork-linking\u002FSKILL.md",[],[],{"skillNo":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":103,"categories":104,"keywords":105,"publishedAt":18},"01M79QZVFA3GTVNZMJ51NESVSP","status-and-learning","Produce a dated, explicitly non-live Linear-to-Notion progress summary, or capture an outcome\u002Flearning after work completes. Use when asked to summarize progress to Notion, report portfolio status, or capture what was learned from a completed piece of work — not to disposition open follow-up items (see open-item-management for that). Never runs automatically or continuously — every summary is a deliberate, approved snapshot, not a live sync.","https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002Fplugins\u002Fworkmanagement-kit\u002Fskills\u002Fstatus-and-learning\u002FSKILL.md",[],[],{"skillNo":107,"name":108,"nameZh":6,"description":109,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":110,"categories":111,"keywords":112,"publishedAt":18},"01M79QZNSKSABK7X2PT93X8QT5","repository-gates","Discover, delegate, record, and invalidate this repository's actual Git\u002FGitHub lifecycle gates (pre-commit, pre-push, PR required checks, Codex delta review, review requirements, merge rights) and the repository-policy provider profile that maps governed operations to git-kit. Use when asked what gates apply to this repository, to discover repository policy before starting Wave 2 work, or to check whether prior gate evidence is still valid after new commits. Discovery and delegation only — never executes a gate's own internal logic, and never invents a gate the repository doesn't actually define (in particular, never assumes a \"Review Changes\" gate exists).","https:\u002F\u002Fgithub.com\u002FAndreHahm\u002Fandres-cc-marketplace\u002Fblob\u002Fmain\u002Fplugins\u002Fworkmanagement-kit\u002Fskills\u002Frepository-gates\u002FSKILL.md",[],[]]