[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6JSM210VV6ES1MAAAS9HPBA":3,"skill-same:01M6JSM210VV6ES1MAAAS9HPBA":90},{"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":27,"publishedAt":48,"linkOnly":49,"markdownContent":50,"markdownBody":51,"frontMatter":52,"refName":70,"commitSha":71,"skillDirectory":72,"forksCount":13,"repositoryHtmlUrl":73,"repositoryHomepageUrl":6,"ownerHtmlUrl":74,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":75,"install":76,"resources":89},"01M6JSM210VV6ES1MAAAS9HPBA","agent-launcher-orchestrator",null,"Use when a user wants to build, launch, grade, or schedule a Claude Managed Agent (CMA) in their own Anthropic account — \"build me an agent\", \"launch this as a managed agent\", \"run this on a schedule\", \"grade my agent against a rubric\", \"set up a nightly worker\". Reads the per-session goal (.\u002Fmy-agent\u002Fgoal.json), routes deterministically to one of five phase sub-skills (interview → stage-launch → grade-iterate → run-without-you → wrap-up) via goal_router.py, and compiles the goal+phase into an execution shape (single-pass workflow \u002F bounded grade→iterate loop \u002F recurring cron deployment loop) via loop_compiler.py. Forks context so heavy intake (build sheets, payloads, eval cases) stays out of the parent thread. All launches are emitted as BYOK curl the user runs with their own key; no tool makes API calls. Inspired by anthropics\u002Flaunch-your-agent (Apache-2.0). Distinct from engineering\u002Fagent-harness (generic domain loop) and engineering\u002Fwrite-a-skill (authors Claude Code skills, not CMAs).","JiayiV\u002Fquestion_mixing","JiayiV","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F54466889?v=4","01M6JFJYYMRZ3QRGP9339JNQ2K","01M6JFJZ35KB0QVT50NAHPJA7G",0,"MIT","https:\u002F\u002Fgithub.com\u002FJiayiV\u002Fquestion_mixing\u002Fblob\u002Fmain\u002Fskills\u002Fpublic-02-claude-skills-main-agent-launcher-skills-agent-launcher-orchestrator-skill-486c6542\u002FSKILL.md",[17,22],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",{"categoryNo":23,"categoryName":24,"taxonomyType":25,"taxonomyTypeDesc":26},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47],"launcher","agent launcher orchestrator","orchestrator","claude-managed-agents","you","worker","me","opencode","up","agent-harness","router","cli","eval","schedule","loop","api","SET","apache","code","workflow","2026-11-10 10:12:30",false,"---\nname: agent-launcher-orchestrator\ndescription: Use when a user wants to build, launch, grade, or schedule a Claude Managed Agent (CMA) in their own Anthropic account — \"build me an agent\", \"launch this as a managed agent\", \"run this on a schedule\", \"grade my agent against a rubric\", \"set up a nightly worker\". Reads the per-session goal (.\u002Fmy-agent\u002Fgoal.json), routes deterministically to one of five phase sub-skills (interview → stage-launch → grade-iterate → run-without-you → wrap-up) via goal_router.py, and compiles the goal+phase into an execution shape (single-pass workflow \u002F bounded grade→iterate loop \u002F recurring cron deployment loop) via loop_compiler.py. Forks context so heavy intake (build sheets, payloads, eval cases) stays out of the parent thread. All launches are emitted as BYOK curl the user runs with their own key; no tool makes API calls. Inspired by anthropics\u002Flaunch-your-agent (Apache-2.0). Distinct from engineering\u002Fagent-harness (generic domain loop) and engineering\u002Fwrite-a-skill (authors Claude Code skills, not CMAs).\ncontext: fork\nversion: 2.11.2\nauthor: Alireza Rezvani\nlicense: MIT\ntags: [claude-managed-agents, cma, agent, launch, orchestrator, session-goal, loop, workflow, cron, outcome, byok]\ncompatible_tools: [claude-code, codex-cli, cursor, antigravity, opencode, gemini-cli]\n---\n\n# agent-launcher — Domain Orchestrator\n\nEvery session starts with a **goal** — one sentence for one CMA. This orchestrator\nreads that goal, routes to the right phase, and compiles the goal into a **loop or\na workflow**. Heavy intake stays in the forked context; the parent gets a digest.\n\nInspired by Anthropic's `launch-your-agent` reference skill (Apache-2.0). This is\nan independent re-implementation; CMA semantics come from\n[`..\u002F..\u002Freferences\u002Fcma-primitives.md`](..\u002F..\u002Freferences\u002Fcma-primitives.md).\n\n## The through-line: the session goal\n\nState lives at `.\u002Fmy-agent\u002Fgoal.json` (the user's folder). Manage it with\n`goal_state.py` (init \u002F set \u002F status \u002F advance) — it also backs the `\u002Fcs:goal`\ncommand and the opt-in `SessionStart` hook. The goal's `phase` selects the lane;\nthe phase + recurrence selects the loop shape.\n\n## Routing (deterministic)\n\nRun the router, then act on its exit code:\n\n```bash\npython3 scripts\u002Fgoal_router.py --out-dir .\u002Fmy-agent\n# exit 0 ROUTE  -> fork to the named phase sub-skill\n# exit 3 ASK    -> ask the one printed forcing question, then re-route\n# exit 4 REFUSE -> goal too vague; get one sentence, then re-route\n```\n\n| Lane (phase) | Sub-skill | Loop\u002Fworkflow |\n|---|---|---|\n| interview | `interview` | single-pass workflow |\n| stage-launch | `stage-launch` | single-pass workflow |\n| grade-iterate | `grade-iterate` | **bounded grade→iterate loop** |\n| run-without-you | `run-without-you` | **recurring cron deployment loop** |\n| wrap-up | `wrap-up` | — |\n\n## Compile the loop\n\n```bash\npython3 scripts\u002Floop_compiler.py \\\n  --out-dir .\u002Fmy-agent --max-iterations 5 --cron \"0 9 * * *\" --timezone Europe\u002FBerlin --nest-outcome\n```\n\n`loop_compiler.py` emits `plan.v1`: `single-pass`, `grade-iterate` (always with a\n`max_iterations` cap 1..20), or `cron-loop` (optionally nesting a self-grading\noutcome per firing). See [`..\u002F..\u002Freferences\u002Floops-and-workflows.md`](..\u002F..\u002Freferences\u002Floops-and-workflows.md).\n\n## Pre-flight gates (hard refusals)\n\n1. **No goal set.** If `goal.json` is missing, run\n   `goal_state.py init --goal \"...\"` first. The orchestrator does not guess a goal.\n2. **Goal too vague.** Router exit 4 — get one sentence naming the one job before\n   routing. Never route on under-3-word goals.\n3. **Never make API calls.** Emit BYOK curl; the user runs it with their own\n   `$ANTHROPIC_API_KEY`. No script in this plugin touches the network.\n4. **Never print the key.** Launch scripts read the key from the environment.\n\n## Hand-off contract\n\nAfter routing, fork to the sub-skill with: the goal string, `agent_name`,\n`out_dir` (`.\u002Fmy-agent`), and the compiled `plan.v1`. When the sub-skill returns,\n`goal_state.py advance` moves the phase and the parent gets a ≤100-word digest\n(phase done, artifact paths, loop shape, one next step).\n\n## Forcing-question library (walk one at a time; recommend + cite)\n\n1. **\"What one job should this agent do end-to-end?\"** — *Recommend:* the single\n   most repeated task. *Cite:* interview-to-config.md (six intake slots). Refuse to\n   route a two-job goal; split into two `.\u002Fmy-agent-*\u002F` folders.\n2. **\"What kicks it off — you ask it, an event, or a schedule?\"** — *Recommend:*\n   on-demand for v0, schedule as the Phase-4 upgrade. *Cite:* loops-and-workflows.md.\n3. **\"How would you grade a good run?\"** — *Recommend:* 3–5 rubric lines grounded\n   in the output. *Cite:* cma-primitives.md (outcomes; rubric required).\n4. **\"Is a real integration ready, or do we mock it in v0?\"** — *Recommend:* mock\n   with a schema-true custom tool; wire the MCP server as v1. *Cite:* interview-to-config.md.\n5. **\"Should run #10 be smarter than run #1?\"** — *Recommend:* attach a memory\n   store only if yes; else skip it. *Cite:* cma-primitives.md (memory limits + injection risk).\n\n## Tools\n\n- `scripts\u002Fgoal_state.py` — own `goal.json` (init\u002Fset\u002Fstatus\u002Fadvance).\n- `scripts\u002Fgoal_router.py` — goal → lane (exit 0 route \u002F 3 ask \u002F 4 refuse).\n- `scripts\u002Floop_compiler.py` — goal+phase → `plan.v1` execution shape.\n","# agent-launcher — Domain Orchestrator\n\nEvery session starts with a **goal** — one sentence for one CMA. This orchestrator\nreads that goal, routes to the right phase, and compiles the goal into a **loop or\na workflow**. Heavy intake stays in the forked context; the parent gets a digest.\n\nInspired by Anthropic's `launch-your-agent` reference skill (Apache-2.0). This is\nan independent re-implementation; CMA semantics come from\n[`..\u002F..\u002Freferences\u002Fcma-primitives.md`](..\u002F..\u002Freferences\u002Fcma-primitives.md).\n\n## The through-line: the session goal\n\nState lives at `.\u002Fmy-agent\u002Fgoal.json` (the user's folder). Manage it with\n`goal_state.py` (init \u002F set \u002F status \u002F advance) — it also backs the `\u002Fcs:goal`\ncommand and the opt-in `SessionStart` hook. The goal's `phase` selects the lane;\nthe phase + recurrence selects the loop shape.\n\n## Routing (deterministic)\n\nRun the router, then act on its exit code:\n\n```bash\npython3 scripts\u002Fgoal_router.py --out-dir .\u002Fmy-agent\n# exit 0 ROUTE  -> fork to the named phase sub-skill\n# exit 3 ASK    -> ask the one printed forcing question, then re-route\n# exit 4 REFUSE -> goal too vague; get one sentence, then re-route\n```\n\n| Lane (phase) | Sub-skill | Loop\u002Fworkflow |\n|---|---|---|\n| interview | `interview` | single-pass workflow |\n| stage-launch | `stage-launch` | single-pass workflow |\n| grade-iterate | `grade-iterate` | **bounded grade→iterate loop** |\n| run-without-you | `run-without-you` | **recurring cron deployment loop** |\n| wrap-up | `wrap-up` | — |\n\n## Compile the loop\n\n```bash\npython3 scripts\u002Floop_compiler.py \\\n  --out-dir .\u002Fmy-agent --max-iterations 5 --cron \"0 9 * * *\" --timezone Europe\u002FBerlin --nest-outcome\n```\n\n`loop_compiler.py` emits `plan.v1`: `single-pass`, `grade-iterate` (always with a\n`max_iterations` cap 1..20), or `cron-loop` (optionally nesting a self-grading\noutcome per firing). See [`..\u002F..\u002Freferences\u002Floops-and-workflows.md`](..\u002F..\u002Freferences\u002Floops-and-workflows.md).\n\n## Pre-flight gates (hard refusals)\n\n1. **No goal set.** If `goal.json` is missing, run\n   `goal_state.py init --goal \"...\"` first. The orchestrator does not guess a goal.\n2. **Goal too vague.** Router exit 4 — get one sentence naming the one job before\n   routing. Never route on under-3-word goals.\n3. **Never make API calls.** Emit BYOK curl; the user runs it with their own\n   `$ANTHROPIC_API_KEY`. No script in this plugin touches the network.\n4. **Never print the key.** Launch scripts read the key from the environment.\n\n## Hand-off contract\n\nAfter routing, fork to the sub-skill with: the goal string, `agent_name`,\n`out_dir` (`.\u002Fmy-agent`), and the compiled `plan.v1`. When the sub-skill returns,\n`goal_state.py advance` moves the phase and the parent gets a ≤100-word digest\n(phase done, artifact paths, loop shape, one next step).\n\n## Forcing-question library (walk one at a time; recommend + cite)\n\n1. **\"What one job should this agent do end-to-end?\"** — *Recommend:* the single\n   most repeated task. *Cite:* interview-to-config.md (six intake slots). Refuse to\n   route a two-job goal; split into two `.\u002Fmy-agent-*\u002F` folders.\n2. **\"What kicks it off — you ask it, an event, or a schedule?\"** — *Recommend:*\n   on-demand for v0, schedule as the Phase-4 upgrade. *Cite:* loops-and-workflows.md.\n3. **\"How would you grade a good run?\"** — *Recommend:* 3–5 rubric lines grounded\n   in the output. *Cite:* cma-primitives.md (outcomes; rubric required).\n4. **\"Is a real integration ready, or do we mock it in v0?\"** — *Recommend:* mock\n   with a schema-true custom tool; wire the MCP server as v1. *Cite:* interview-to-config.md.\n5. **\"Should run #10 be smarter than run #1?\"** — *Recommend:* attach a memory\n   store only if yes; else skip it. *Cite:* cma-primitives.md (memory limits + injection risk).\n\n## Tools\n\n- `scripts\u002Fgoal_state.py` — own `goal.json` (init\u002Fset\u002Fstatus\u002Fadvance).\n- `scripts\u002Fgoal_router.py` — goal → lane (exit 0 route \u002F 3 ask \u002F 4 refuse).\n- `scripts\u002Floop_compiler.py` — goal+phase → `plan.v1` execution shape.\n",{"name":5,"tags":53,"author":61,"context":62,"license":14,"version":63,"description":7,"compatible_tools":64},[31,54,55,56,30,57,42,47,58,59,60],"cma","agent","launch","session-goal","cron","outcome","byok","Alireza Rezvani","fork","2.11.2",[65,66,67,68,35,69],"claude-code","codex-cli","cursor","antigravity","gemini-cli","main","068c2486ccc0f8d200a3db4d21ec137064eda3b4","skills\u002Fpublic-02-claude-skills-main-agent-launcher-skills-agent-launcher-orchestrator-skill-486c6542","https:\u002F\u002Fgithub.com\u002FJiayiV\u002Fquestion_mixing","https:\u002F\u002Fgithub.com\u002FJiayiV","2026-08-25 16:30:52",{"source":8,"skillName":5,"commands":77,"prompt":87,"zipAvailable":88},[78,81,84],{"tool":79,"command":80},"npx","npx skills add JiayiV\u002Fquestion_mixing --skill agent-launcher-orchestrator",{"tool":82,"command":83},"bunx","bunx skills add JiayiV\u002Fquestion_mixing --skill agent-launcher-orchestrator",{"tool":85,"command":86},"pnpm","pnpm dlx skills add JiayiV\u002Fquestion_mixing --skill agent-launcher-orchestrator","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FJiayiV\u002Fquestion_mixing\u002Fblob\u002Fmain\u002Fskills\u002Fpublic-02-claude-skills-main-agent-launcher-skills-agent-launcher-orchestrator-skill-486c6542\u002FSKILL.md",true,[],[91,116,142,171,198,230],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":48},"01M6NN9PJS9NHWPDG4QP93BP8F","webmcpify","Make a web app agent-ready — propose a WebMCP tool manifest, integrate, verify in a real browser, heal; unrelated code stays untouched. Use for \"webmcpify\", \"add WebMCP\", or \"expose app actions to AI agents\".","https:\u002F\u002Fgithub.com\u002FJiayiV\u002Fquestion_mixing\u002Fblob\u002Fmain\u002Fskills\u002Fwebmcpify\u002FSKILL.md",[],[93,98,99,100,101,46,102,103,104,105,106,107,108,109,110,111,112,113,114,115],"github","ai agents","web","notes","browser","https","actions","manifest","inventory","verify","tool","use","ready","unrelated","app","scope","for","webmcp",{"skillNo":117,"name":118,"nameZh":6,"description":119,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":120,"licenseFileUrl":6,"sourceHtmlUrl":121,"categories":122,"keywords":123,"publishedAt":48},"01M6NN6KZE5RBN6NGHN2WXV1RE","vardoger-analyze","Use when the user asks to personalize the GitHub Copilot CLI assistant, adapt Copilot to their style, use vardoger, or analyze their Copilot CLI conversation history. Reads the local session directory at `~\u002F.copilot\u002Fsession-state\u002F`, extracts recurring preferences and conventions, and writes a fenced personalization block into `~\u002F.copilot\u002Fcopilot-instructions.md`. Runs entirely on the user's machine via the local `vardoger` CLI (`pipx install vardoger`); no network calls and no uploads. Triggers: 'personalize my copilot', 'analyze my copilot history', 'tailor copilot to me', 'run vardoger', 'update my copilot instructions from history', 'make copilot learn my style'.","Apache-2.0","https:\u002F\u002Fgithub.com\u002FJiayiV\u002Fquestion_mixing\u002Fblob\u002Fmain\u002Fskills\u002Fvardoger-analyze\u002FSKILL.md",[],[124,125,126,39,127,128,129,130,131,132,133,134,135,34,136,137,138,139,140,141],"vardoger analyze","analyze","directory","github copilot","copilot","run","update","learn","network","my style","install","conversation","local","github copilot cli","pipx","my","at","copilot instructions",{"skillNo":143,"name":144,"nameZh":6,"description":145,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":146,"categories":147,"keywords":152,"publishedAt":48},"01M6NN3CZGQP3XAJ04BM1NNYGT","transloadit-media-processing","Process media files (video, audio, images, documents) using Transloadit. Use when asked to encode video to HLS\u002FMP4, generate thumbnails, resize or watermark images, extract audio, concatenate clips, add subtitles, OCR documents, or run any media processing pipeline. Covers 86+ processing robots for file transformation at scale.","https:\u002F\u002Fgithub.com\u002FJiayiV\u002Fquestion_mixing\u002Fblob\u002Fmain\u002Fskills\u002Ftransloadit-media-processing\u002FSKILL.md",[148,151],{"categoryNo":149,"categoryName":150,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000016","图像与多媒体",{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},[153,154,155,156,157,158,159,160,161,162,163,164,165,109,166,167,168,169,114,170],"transloadit media processing","Media","media processing","processing","MCP","mcp server","documents","subtitles","pipeline","ocr","Video","Audio","mp4","images","files","transformation","free","hls",{"skillNo":172,"name":173,"nameZh":6,"description":174,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":175,"categories":176,"keywords":180,"publishedAt":48},"01M6NN28E92F9GH4WFYQ2H1Y15","terraform-azurerm-set-diff-analyzer","Analyze Terraform plan JSON output for AzureRM Provider to distinguish between false-positive diffs (order-only changes in Set-type attributes) and actual resource changes. Use when reviewing terraform plan output for Azure resources like Application Gateway, Load Balancer, Firewall, Front Door, NSG, and other resources with Set-type attributes that cause spurious diffs due to internal ordering changes.","https:\u002F\u002Fgithub.com\u002FJiayiV\u002Fquestion_mixing\u002Fblob\u002Fmain\u002Fskills\u002Fterraform-azurerm-set-diff-analyzer\u002FSKILL.md",[177],{"categoryNo":178,"categoryName":179,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000011","运维与部署",[181,44,182,183,184,185,186,187,188,109,189,190,191,192,193,194,195,196,197,114],"terraform","terraform azurerm set diff analyzer","diff","analyzer","diffs","azure","gateway","firewall","only","resource changes","provider","resources","load balancer","attributes","front","load","balancer",{"skillNo":199,"name":200,"nameZh":6,"description":201,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":202,"categories":203,"keywords":213,"publishedAt":48},"01M6NN1QMH2TDYWTH3XY3N5XPV","technical-job-search","Use this skill when a software engineer asks for help with job search tasks: parsing or analyzing a job description, tailoring a CV\u002Fresume, writing a cover letter, evaluating a job offer, or drafting a post-interview follow-up email. Do not activate for general career advice unrelated to an active job search action.","https:\u002F\u002Fgithub.com\u002FJiayiV\u002Fquestion_mixing\u002Fblob\u002Fmain\u002Fskills\u002Ftechnical-job-search\u002FSKILL.md",[204,207,210],{"categoryNo":205,"categoryName":206,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000018","写作与文案",{"categoryNo":208,"categoryName":209,"taxonomyType":25,"taxonomyTypeDesc":26},"01JBSEEDCAT000000000000036","内容创作者",{"categoryNo":211,"categoryName":212,"taxonomyType":25,"taxonomyTypeDesc":26},"01JBSEEDCAT000000000000038","人力资源",[214,215,216,217,218,219,220,221,125,222,109,36,223,224,225,226,227,228,114,229],"search","Job Search","technical job search","technical","writing","tasks","email","parsing","interview","engineer","general","activate","do","task","cv","letter",{"skillNo":231,"name":232,"nameZh":6,"description":233,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":234,"categories":235,"keywords":236,"publishedAt":48},"01M6NMZFA67XYS1GS09WT42235","steno-mode","Shorthand-first response compression that cuts ~40% of response tokens while preserving technical precision and exact literals. Use when the user says \"steno mode\", \"shorthand mode\", \"compressed responses\", \"token reduction\", \"brief structured output\", or invokes \u002Fsteno. Supports four compression levels: lite, brief, court, machine. Do not trigger for requests needing polished prose such as onboarding\u002Ftutorial content, stakeholder or customer-facing copy, or teaching-focused explanations.","https:\u002F\u002Fgithub.com\u002FJiayiV\u002Fquestion_mixing\u002Fblob\u002Fmain\u002Fskills\u002Fsteno-mode\u002FSKILL.md",[],[237,238,239,240,241,217,242,109,226,243,244,245,246,247,248,249,250,251,114,252],"steno mode","mode","customer","brief","first","tutorial","teaching","tokens","as","trigger","structured output","machine","output","court","requests","response compression"]