[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M77GH232WEVXP3KJBVWMBCYY":3,"skill-same:01M77GH232WEVXP3KJBVWMBCYY":47},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":6,"ownerDescription":31,"updatedAt":32,"install":33,"resources":46},"01M77GH232WEVXP3KJBVWMBCYY","gateway-restart-resume",null,"Use when an OpenClaw gateway-backed agent is asked through Discord, Slack, Matrix, SMS, web chat, CLI relay, ACP, cron, or another channel to restart, stop, reload, update, or repair the gateway and needs a durable post-restart callback, same-thread reply, recovery status, or lost follow-up prevention.","Lord1Egypt\u002Fawesome-skill-forge","Lord1Egypt","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F99746359?v=4","01M7592YDNT0XK9WJ0T2MQVPDN","Mohamed Mounir Ali AL Keshita","01M75GRFNHC8VZNWEB8DKDE433",3,"MIT","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fcommunity\u002Fclawhub\u002Fg\u002Fgateway-restart-resume\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: gateway-restart-resume\ndescription: Use when an OpenClaw gateway-backed agent is asked through Discord, Slack, Matrix, SMS, web chat, CLI relay, ACP, cron, or another channel to restart, stop, reload, update, or repair the gateway and needs a durable post-restart callback, same-thread reply, recovery status, or lost follow-up prevention.\n---\n\n# OpenClaw Restart Resume\n\n## Overview\n\nGateway restart is a conversation-continuity problem. An in-gateway OpenClaw agent that restarts the gateway kills the runtime that would normally send the follow-up reply. Before taking any gateway-killing action, create a durable resume path that can verify recovery and report back through the same medium, conversation, channel, thread, or user route.\n\n## Runtime Tool Preflight\n\nBefore any in-gateway restart attempt, verify that the current runtime exposes a durable control path:\n\n- `cron` or an equivalent durable task tool for the resume callback.\n- `gateway` or `exec` for the restart action.\n- A delivery path for the original medium, such as cron `announce` delivery, the original session route, or an approved channel send mechanism.\n\nIf any required path is unavailable, do not restart. Reply in the current conversation with a short degraded status:\n\n```text\nI did not restart the gateway because this runtime cannot create a durable post-restart callback first.\nNeeded before retry: cron\u002Ftask callback access plus gateway\u002Fexec restart access.\n```\n\nNever replace this preflight with best-effort memory, a note in the transcript, or an instruction to yourself. The callback must be durable outside the current in-gateway turn.\n\n## Pre-Restart Gate\n\nBefore any restart, explicitly satisfy every item:\n\n- Durable callback created outside the current in-gateway turn.\n- Original reply route captured with no secrets or raw message bodies.\n- Short pre-restart acknowledgement sent to the user.\n- Restart action explicitly requested or approved.\n- Callback includes verification steps, bounded retry budget, and failure reporting.\n\nIf any item is missing, do not restart. Report the missing prerequisite in the current conversation.\n\n## When to Use\n\nUse this skill when:\n\n- A user asks through Discord, Slack, Matrix, SMS, email, web chat, CLI relay, ACP, cron, or another OpenClaw-backed medium to restart, stop, reload, update, or repair the gateway.\n- The action may stop the gateway, restart the service manager, reload plugins, or replace the running OpenClaw package.\n- The agent is running inside OpenClaw, or may be running inside OpenClaw.\n- The user expects a post-restart status reply.\n\nDo not use this skill for external shell operators that can survive the restart themselves, except as a checklist for user-facing notifications.\n\n## Restart Resume Protocol\n\n1. Classify execution context.\n   - If running inside OpenClaw, assume the restart will terminate the current turn.\n   - If running outside OpenClaw, still use the notification and verification steps when the request came from a user-facing channel.\n\n2. Confirm permission.\n   - If the user did not explicitly ask for restart\u002Fstop\u002Fupdate, ask first.\n   - Do not restart for vague requests such as \"fix it\" until the restart is explained and approved.\n\n3. Capture reply target.\n   - Record the medium, account\u002Fprofile if relevant, conversation id, channel id, thread id, message id, user id if needed, and agent id.\n   - Prefer the exact original route. If the medium has no thread concept, record the smallest addressable conversation target.\n   - Store only routing metadata needed to reply. Do not store raw message bodies, tokens, env contents, or secrets.\n\n4. Send pre-restart acknowledgement.\n   - Tell the user the gateway is restarting and that a follow-up will be sent after verification.\n   - Keep it short because the gateway may stop immediately afterward.\n\n5. Create durable resume callback before restart.\n   - Prefer a one-shot or short-delay OpenClaw cron job, task, or native gateway callback that survives the gateway restart.\n   - The callback must include the reply target and verification checklist.\n   - If callback creation fails, do not restart. Tell the user the restart was not attempted because no follow-up path exists.\n   - For OpenClaw cron, prefer a one-shot job with delete-after-run enabled and fallback delivery\u002Fannounce pointed at the original medium target.\n   - For cron-specific command shape, read [references\u002Fopenclaw-cron-resume.md](references\u002Fopenclaw-cron-resume.md) only when using OpenClaw cron.\n\n6. Restart only after the callback exists.\n   - Use the narrowest approved restart action.\n   - Avoid broad update\u002Freinstall operations unless the user requested them and the current admin runbook supports them.\n\n7. Resume callback verifies recovery.\n   - Check gateway\u002Fservice status.\n   - Check health endpoint or gateway probe.\n   - Check connectivity or send capability for the original medium when the local OpenClaw version supports it.\n   - Check plugin or task health only when relevant to the restart reason.\n\n8. Send final reply.\n   - Reply in the original medium and conversation\u002Fthread when available.\n   - Use the operator report format below unless the original medium requires a shorter message.\n   - State whether restart completed, what was verified, duration if known, and any remaining warning.\n   - If verification fails, report the failed check and the safest next action.\n   - Translate raw tool results into human status. Never paste raw strings such as `gateway.restart`, `restart ok`, `restart restart ok`, method names, JSON payloads, or internal reason text into the user-facing reply.\n\n## Medium-Agnostic Reply Rules\n\n- Discord\u002FSlack\u002FMatrix: reply in the original thread when available; otherwise reply in the original channel\u002Fconversation.\n- SMS\u002Fchat apps: keep the final message compact and avoid multi-line technical detail unless the user asked for it.\n- Email: include a concise subject-like first line, then the verification report.\n- CLI relay or ACP: return structured status when the caller expects JSON\u002Ftool output; include human-readable summary when routed to a person.\n- Cron or automation-triggered restarts: send the final report to the configured notification target and record the result in durable run history if available.\n- If the original medium is unavailable after restart, use the configured fallback notification target only when policy allows it; otherwise leave a durable ops note for the operator.\n\n## Resume Payload\n\nThe durable callback should carry a minimal payload:\n\n```json\n{\n  \"reason\": \"user-requested gateway restart\",\n  \"agentId\": \"\u003Cagent-id>\",\n  \"reply\": {\n    \"medium\": \"\u003Cdiscord|slack|matrix|sms|email|web|cli|acp|cron|other>\",\n    \"account\": \"\u003Caccount-id-if-needed>\",\n    \"target\": \"\u003Cconversation-or-addressable-target>\",\n    \"channelId\": \"\u003Cchannel-id-if-needed>\",\n    \"threadId\": \"\u003Cthread-id-if-needed>\",\n    \"userId\": \"\u003Cuser-id-if-needed>\",\n    \"messageId\": \"\u003Coriginal-message-id-if-needed>\"\n  },\n  \"verify\": [\n    \"gateway status\",\n    \"health\",\n    \"original medium connectivity\"\n  ],\n  \"createdAt\": \"\u003Ciso-timestamp>\"\n}\n```\n\nSanitize or omit fields that are not needed for routing. Never include secrets, full config, env files, transcript bodies, or private message contents.\n\n## Verification Checklist\n\nUse the local OpenClaw version's native tools when available. Equivalent checks are acceptable.\n\n```bash\nopenclaw gateway status\nopenclaw health\nopenclaw channels status --probe\n```\n\nWhen the restart was part of an update or plugin repair, also verify the relevant narrow checks, for example:\n\n```bash\nopenclaw doctor --non-interactive --no-workspace-suggestions\nopenclaw plugins doctor\nopenclaw tasks audit\n```\n\nDo not claim the restart succeeded unless fresh verification output supports it.\n\n## Failure Handling\n\n- Callback creation fails: do not restart; explain that there is no durable way to report back.\n- Restart command fails before shutdown: report the command failure immediately in the current medium.\n- Gateway returns but the original medium is disconnected: report through any still-available approved fallback route, or leave a durable task\u002Fops note for the operator.\n- Gateway does not return: retry verification up to 3 total attempts over about 5 minutes, then record the failure in durable state and stop.\n- Auth\u002Ftoken errors appear: do not read or paste secrets. Report the config path\u002Fkey names only and ask for explicit operator direction.\n\n## Common Mistakes\n\n- Restarting first and trying to schedule the follow-up afterward.\n- Treating \"gateway command returned\" as proof that the original medium recovered.\n- Storing raw messages or secret-bearing config in the resume payload.\n- Creating a repeating cron when a one-shot callback is enough.\n- Retrying forever after a failed restart.\n- Sending the final reply to a default destination instead of the original medium\u002Fconversation\u002Fthread.\n\n## Output Message Format\n\nPrefer this operator report for human-facing media that can handle multi-line messages.\n\nOutput must be written as a clean operator status, not as a tool transcript.\n\nDo not include:\n\n- Raw tool names or method names, such as `gateway.restart`.\n- Duplicated\u002Fgenerated phrases, such as \"restart restart ok\".\n- Internal reason payloads unless the user explicitly asks for them.\n- \"Recommended follow-up: run openclaw doctor...\" after a successful restart. Only suggest a terminal command when verification failed or is genuinely blocked.\n\nBefore restart:\n\n```text\nRestarting the OpenClaw gateway now.\nI will reply here after it comes back.\n```\n\nAfter success:\n\n```text\nGateway restart complete.\n\nVerified:\n- Gateway: back online\n- Discord: reply path restored\n- Resume: this follow-up reached the original thread\n\nDuration: \u003Cduration or \"not recorded\">\nWarnings: \u003Cnone or one short warning>\n```\n\nAfter degraded recovery:\n\n```text\nGateway restart completed, but verification is degraded.\n\nPassed:\n- \u003Cpassed check>\n\nNeeds attention:\n- \u003Cfailed or warning check>\n\nSafest next action: \u003Cnext step>.\n```\n\nAfter failure:\n\n```text\nGateway restart did not verify cleanly.\n\nFailed check: \u003Ccheck>\nLast known state: \u003Cstate>\nSafest next action: \u003Cnext step>\n```\n","# OpenClaw Restart Resume\n\n## Overview\n\nGateway restart is a conversation-continuity problem. An in-gateway OpenClaw agent that restarts the gateway kills the runtime that would normally send the follow-up reply. Before taking any gateway-killing action, create a durable resume path that can verify recovery and report back through the same medium, conversation, channel, thread, or user route.\n\n## Runtime Tool Preflight\n\nBefore any in-gateway restart attempt, verify that the current runtime exposes a durable control path:\n\n- `cron` or an equivalent durable task tool for the resume callback.\n- `gateway` or `exec` for the restart action.\n- A delivery path for the original medium, such as cron `announce` delivery, the original session route, or an approved channel send mechanism.\n\nIf any required path is unavailable, do not restart. Reply in the current conversation with a short degraded status:\n\n```text\nI did not restart the gateway because this runtime cannot create a durable post-restart callback first.\nNeeded before retry: cron\u002Ftask callback access plus gateway\u002Fexec restart access.\n```\n\nNever replace this preflight with best-effort memory, a note in the transcript, or an instruction to yourself. The callback must be durable outside the current in-gateway turn.\n\n## Pre-Restart Gate\n\nBefore any restart, explicitly satisfy every item:\n\n- Durable callback created outside the current in-gateway turn.\n- Original reply route captured with no secrets or raw message bodies.\n- Short pre-restart acknowledgement sent to the user.\n- Restart action explicitly requested or approved.\n- Callback includes verification steps, bounded retry budget, and failure reporting.\n\nIf any item is missing, do not restart. Report the missing prerequisite in the current conversation.\n\n## When to Use\n\nUse this skill when:\n\n- A user asks through Discord, Slack, Matrix, SMS, email, web chat, CLI relay, ACP, cron, or another OpenClaw-backed medium to restart, stop, reload, update, or repair the gateway.\n- The action may stop the gateway, restart the service manager, reload plugins, or replace the running OpenClaw package.\n- The agent is running inside OpenClaw, or may be running inside OpenClaw.\n- The user expects a post-restart status reply.\n\nDo not use this skill for external shell operators that can survive the restart themselves, except as a checklist for user-facing notifications.\n\n## Restart Resume Protocol\n\n1. Classify execution context.\n   - If running inside OpenClaw, assume the restart will terminate the current turn.\n   - If running outside OpenClaw, still use the notification and verification steps when the request came from a user-facing channel.\n\n2. Confirm permission.\n   - If the user did not explicitly ask for restart\u002Fstop\u002Fupdate, ask first.\n   - Do not restart for vague requests such as \"fix it\" until the restart is explained and approved.\n\n3. Capture reply target.\n   - Record the medium, account\u002Fprofile if relevant, conversation id, channel id, thread id, message id, user id if needed, and agent id.\n   - Prefer the exact original route. If the medium has no thread concept, record the smallest addressable conversation target.\n   - Store only routing metadata needed to reply. Do not store raw message bodies, tokens, env contents, or secrets.\n\n4. Send pre-restart acknowledgement.\n   - Tell the user the gateway is restarting and that a follow-up will be sent after verification.\n   - Keep it short because the gateway may stop immediately afterward.\n\n5. Create durable resume callback before restart.\n   - Prefer a one-shot or short-delay OpenClaw cron job, task, or native gateway callback that survives the gateway restart.\n   - The callback must include the reply target and verification checklist.\n   - If callback creation fails, do not restart. Tell the user the restart was not attempted because no follow-up path exists.\n   - For OpenClaw cron, prefer a one-shot job with delete-after-run enabled and fallback delivery\u002Fannounce pointed at the original medium target.\n   - For cron-specific command shape, read [references\u002Fopenclaw-cron-resume.md](references\u002Fopenclaw-cron-resume.md) only when using OpenClaw cron.\n\n6. Restart only after the callback exists.\n   - Use the narrowest approved restart action.\n   - Avoid broad update\u002Freinstall operations unless the user requested them and the current admin runbook supports them.\n\n7. Resume callback verifies recovery.\n   - Check gateway\u002Fservice status.\n   - Check health endpoint or gateway probe.\n   - Check connectivity or send capability for the original medium when the local OpenClaw version supports it.\n   - Check plugin or task health only when relevant to the restart reason.\n\n8. Send final reply.\n   - Reply in the original medium and conversation\u002Fthread when available.\n   - Use the operator report format below unless the original medium requires a shorter message.\n   - State whether restart completed, what was verified, duration if known, and any remaining warning.\n   - If verification fails, report the failed check and the safest next action.\n   - Translate raw tool results into human status. Never paste raw strings such as `gateway.restart`, `restart ok`, `restart restart ok`, method names, JSON payloads, or internal reason text into the user-facing reply.\n\n## Medium-Agnostic Reply Rules\n\n- Discord\u002FSlack\u002FMatrix: reply in the original thread when available; otherwise reply in the original channel\u002Fconversation.\n- SMS\u002Fchat apps: keep the final message compact and avoid multi-line technical detail unless the user asked for it.\n- Email: include a concise subject-like first line, then the verification report.\n- CLI relay or ACP: return structured status when the caller expects JSON\u002Ftool output; include human-readable summary when routed to a person.\n- Cron or automation-triggered restarts: send the final report to the configured notification target and record the result in durable run history if available.\n- If the original medium is unavailable after restart, use the configured fallback notification target only when policy allows it; otherwise leave a durable ops note for the operator.\n\n## Resume Payload\n\nThe durable callback should carry a minimal payload:\n\n```json\n{\n  \"reason\": \"user-requested gateway restart\",\n  \"agentId\": \"\u003Cagent-id>\",\n  \"reply\": {\n    \"medium\": \"\u003Cdiscord|slack|matrix|sms|email|web|cli|acp|cron|other>\",\n    \"account\": \"\u003Caccount-id-if-needed>\",\n    \"target\": \"\u003Cconversation-or-addressable-target>\",\n    \"channelId\": \"\u003Cchannel-id-if-needed>\",\n    \"threadId\": \"\u003Cthread-id-if-needed>\",\n    \"userId\": \"\u003Cuser-id-if-needed>\",\n    \"messageId\": \"\u003Coriginal-message-id-if-needed>\"\n  },\n  \"verify\": [\n    \"gateway status\",\n    \"health\",\n    \"original medium connectivity\"\n  ],\n  \"createdAt\": \"\u003Ciso-timestamp>\"\n}\n```\n\nSanitize or omit fields that are not needed for routing. Never include secrets, full config, env files, transcript bodies, or private message contents.\n\n## Verification Checklist\n\nUse the local OpenClaw version's native tools when available. Equivalent checks are acceptable.\n\n```bash\nopenclaw gateway status\nopenclaw health\nopenclaw channels status --probe\n```\n\nWhen the restart was part of an update or plugin repair, also verify the relevant narrow checks, for example:\n\n```bash\nopenclaw doctor --non-interactive --no-workspace-suggestions\nopenclaw plugins doctor\nopenclaw tasks audit\n```\n\nDo not claim the restart succeeded unless fresh verification output supports it.\n\n## Failure Handling\n\n- Callback creation fails: do not restart; explain that there is no durable way to report back.\n- Restart command fails before shutdown: report the command failure immediately in the current medium.\n- Gateway returns but the original medium is disconnected: report through any still-available approved fallback route, or leave a durable task\u002Fops note for the operator.\n- Gateway does not return: retry verification up to 3 total attempts over about 5 minutes, then record the failure in durable state and stop.\n- Auth\u002Ftoken errors appear: do not read or paste secrets. Report the config path\u002Fkey names only and ask for explicit operator direction.\n\n## Common Mistakes\n\n- Restarting first and trying to schedule the follow-up afterward.\n- Treating \"gateway command returned\" as proof that the original medium recovered.\n- Storing raw messages or secret-bearing config in the resume payload.\n- Creating a repeating cron when a one-shot callback is enough.\n- Retrying forever after a failed restart.\n- Sending the final reply to a default destination instead of the original medium\u002Fconversation\u002Fthread.\n\n## Output Message Format\n\nPrefer this operator report for human-facing media that can handle multi-line messages.\n\nOutput must be written as a clean operator status, not as a tool transcript.\n\nDo not include:\n\n- Raw tool names or method names, such as `gateway.restart`.\n- Duplicated\u002Fgenerated phrases, such as \"restart restart ok\".\n- Internal reason payloads unless the user explicitly asks for them.\n- \"Recommended follow-up: run openclaw doctor...\" after a successful restart. Only suggest a terminal command when verification failed or is genuinely blocked.\n\nBefore restart:\n\n```text\nRestarting the OpenClaw gateway now.\nI will reply here after it comes back.\n```\n\nAfter success:\n\n```text\nGateway restart complete.\n\nVerified:\n- Gateway: back online\n- Discord: reply path restored\n- Resume: this follow-up reached the original thread\n\nDuration: \u003Cduration or \"not recorded\">\nWarnings: \u003Cnone or one short warning>\n```\n\nAfter degraded recovery:\n\n```text\nGateway restart completed, but verification is degraded.\n\nPassed:\n- \u003Cpassed check>\n\nNeeds attention:\n- \u003Cfailed or warning check>\n\nSafest next action: \u003Cnext step>.\n```\n\nAfter failure:\n\n```text\nGateway restart did not verify cleanly.\n\nFailed check: \u003Ccheck>\nLast known state: \u003Cstate>\nSafest next action: \u003Cnext step>\n```\n",{"name":5,"description":7},"master","4aa004da461d820f4ea5499e58eb8d196cf99016","community\u002Fclawhub\u002Fg\u002Fgateway-restart-resume",0,"https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge","https:\u002F\u002Fskill-explorer-rho.vercel.app","https:\u002F\u002Fgithub.com\u002FLord1Egypt","Unbreakable. Unstoppable. Unmatched⚡.\r\n Speaker of machine tongues 🤖  | World-builder of decentralized dreams 🌐 | Defender of the digital realm🛡️.","2026-06-10 08:21:01",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add Lord1Egypt\u002Fawesome-skill-forge --skill gateway-restart-resume",{"tool":39,"command":40},"bunx","bunx skills add Lord1Egypt\u002Fawesome-skill-forge --skill gateway-restart-resume",{"tool":42,"command":43},"pnpm","pnpm dlx skills add Lord1Egypt\u002Fawesome-skill-forge --skill gateway-restart-resume","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fcommunity\u002Fclawhub\u002Fg\u002Fgateway-restart-resume\u002FSKILL.md",true,[],[48,55,62,69,76,83],{"skillNo":49,"name":50,"nameZh":6,"description":51,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":52,"categories":53,"keywords":54,"publishedAt":19},"01M785WKJB99G3STR1NF9NRTJW","youtube-content","YouTube transcripts to summaries, threads, blogs.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fyoutube-content\u002FSKILL.md",[],[],{"skillNo":56,"name":57,"nameZh":6,"description":58,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":59,"categories":60,"keywords":61,"publishedAt":19},"01M785WEFHB0G2S9YA6EPR3H0E","xurl","X\u002FTwitter via xurl CLI: post, search, DM, media, v2 API.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fxurl\u002FSKILL.md",[],[],{"skillNo":63,"name":64,"nameZh":6,"description":65,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":66,"categories":67,"keywords":68,"publishedAt":19},"01M785W98EDN4EFJXJPMQEHCRC","touchdesigner-mcp","Control a running TouchDesigner instance via twozero MCP — create operators, set parameters, wire connections, execute Python, build real-time visuals. 36 native tools.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Ftouchdesigner-mcp\u002FSKILL.md",[],[],{"skillNo":70,"name":71,"nameZh":6,"description":72,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"publishedAt":19},"01M785W5S9FP3QBH2V191E025P","test-driven-development","TDD: enforce RED-GREEN-REFACTOR, tests before code.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Ftest-driven-development\u002FSKILL.md",[],[],{"skillNo":77,"name":78,"nameZh":6,"description":79,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":19},"01M785W2AS08NSSH9YATCV6FJB","teams-meeting-pipeline","Operate the Teams meeting summary pipeline via Hermes CLI — summarize meetings, inspect pipeline status, replay jobs, manage Microsoft Graph subscriptions.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fteams-meeting-pipeline\u002FSKILL.md",[],[],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":19},"01M785VXKW4T34BTP1SCHW1R7F","systematic-debugging","4-phase root cause debugging: understand bugs before fixing.","https:\u002F\u002Fgithub.com\u002FLord1Egypt\u002Fawesome-skill-forge\u002Fblob\u002Fmaster\u002Fskills\u002Fsystematic-debugging\u002FSKILL.md",[],[]]