[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6KKHT4N5VX7JBJ3WHK3PPD6":3,"skill-same:01M6KKHT4N5VX7JBJ3WHK3PPD6":76},{"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":36,"linkOnly":37,"markdownContent":38,"markdownBody":39,"frontMatter":40,"refName":42,"commitSha":43,"skillDirectory":44,"forksCount":45,"repositoryHtmlUrl":46,"repositoryHomepageUrl":6,"ownerHtmlUrl":47,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":48,"install":49,"resources":61},"01M6KKHT4N5VX7JBJ3WHK3PPD6","subagent-driven-development-codex",null,"Execute an implementation plan by dispatching Codex-backed implementer and reviewer agents while Claude keeps spec-compliance control. Codex-backed variant.","mzored\u002Fsuperpowers-cc-to-codex","mzored","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F29492942?v=4","01M6KKGRNSJKPE3WQD8AH2DCA0","Миша","01M6KKGRSPMYR9YEA4Q04ZA8YC",7,"MIT","https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Fskills\u002Fsubagent-driven-development-codex\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35],"codex","subagent driven development codex","driven","subagent","subagent driven development","development","compliance","plan","reviewer","execute","implementer","implementation","control","true","variant","implementation plan","spec","2026-11-10 10:12:30",false,"---\nname: subagent-driven-development-codex\ndescription: Execute an implementation plan by dispatching Codex-backed implementer and reviewer agents while Claude keeps spec-compliance control. Codex-backed variant.\ndisable-model-invocation: true\n---\n\u003C!--\nUpstream source: obra\u002Fsuperpowers skills\u002Fsubagent-driven-development\u002FSKILL.md\nLast synced: 2026-04-03\nDivergence: Codex-backed implementer\u002Freviewer agents replace Claude subagents; adapter-managed resume flow\n-->\n\n# Subagent-Driven Development\n\nExecute plan by calling `codex_implement` per task via the MCP tool, with two-stage review after each: Claude-side spec compliance first, then `codex_review` for code quality.\n\n**Core principle:** Codex implementer per task + two-stage review (spec then quality) = high quality, fast iteration\n\n\u003CHARD-GATE>\nYou MUST dispatch codex_implement for every plan task (Step 3).\nYou cannot implement code yourself, even for simple tasks.\nSpec review requires seeing EXTERNAL implementation. No dispatch = no task completion.\n\u003C\u002FHARD-GATE>\n\n## Checklist\n\nYou MUST create a task for each plan task and complete them in order.\n\n**Setup:**\n\n1. **Read plan** — read the plan file once, extract all tasks with full text and context\n2. **Create TodoWrite** — one entry per plan task\n\n**Per task:**\n\n3. **Capture task base** — before dispatching the implementer, record the current revision:\n\n```bash\nTASK_BASE_SHA=\"$(git rev-parse HEAD)\"\n```\n\nSubstitute the captured TASK_BASE_SHA value as the structured review base for this task, so the reviewer inspects only the task's delta rather than all prior tasks.\n\n4. **Dispatch implementer** — call the `codex_implement` MCP tool.\n   Pass the task description as the `prompt`:\n\n```json\n{\n  \"tool\": \"codex_implement\",\n  \"arguments\": {\n    \"taskId\": \"task-17\",\n    \"prompt\": \"Implement Task 3 from docs\u002Fsuperpowers\u002Fplans\u002F2026-04-03-agent-forwarding.md. Keep the work scoped to agents\u002Fcodex-implementer.md and tests\u002Fprompt-contracts\u002Fexecution-workflows.test.mjs.\",\n    \"workspaceRoot\": \"\u002Fabsolute\u002Fpath\u002Fto\u002Fyour\u002Frepo\"\n  }\n}\n```\n\n5. **Handle status** — see Status Handling below\n6. **Spec compliance review** — Claude reads the actual code and verifies against spec using `spec-review-template.md`; if issues remain, resume the same implementer thread with `codex_resume`:\n\n```json\n{\n  \"tool\": \"codex_resume\",\n  \"arguments\": {\n    \"taskId\": \"task-17\",\n    \"sessionId\": \"019d4f82-58b8-72d3-9212-2e3d3fc69bcb\",\n    \"prompt\": \"Fix these issues:\\n- The agent still lacks tools: Bash.\\n- The prompt-file override is ignored in the TDD path.\",\n    \"workspaceRoot\": \"\u002Fabsolute\u002Fpath\u002Fto\u002Fyour\u002Frepo\"\n  }\n}\n```\n\n7. **Code quality review** — call `codex_review` only after spec compliance passes:\n\n```json\n{\n  \"tool\": \"codex_review\",\n  \"arguments\": {\n    \"taskId\": \"task-17-review\",\n    \"reviewStyle\": \"structured\",\n    \"scope\": { \"kind\": \"base\", \"base\": \"\u003Ccaptured TASK_BASE_SHA value>\" },\n    \"prompt\": \"Review the implementation of Task 4 from docs\u002Fsuperpowers\u002Fplans\u002F2026-04-03-agent-forwarding.md. Focus on regressions, incorrect routing, and missing test coverage.\",\n    \"workspaceRoot\": \"\u002Fabsolute\u002Fpath\u002Fto\u002Fyour\u002Frepo\"\n  }\n}\n```\n\n8. **Mark complete** — only after both gates pass\n\n**After all tasks:**\n\n9. **Final review** — call `codex_review` for the entire implementation with a new task ID and the final base SHA.\n10. **Finish branch** — wrap up the development branch\n\n## Status Handling\n\nHandle the implementer's reported status:\n\n| Status | Action |\n|---|---|\n| **DONE** | Proceed to spec compliance review |\n| **DONE_WITH_CONCERNS** | Read concerns; if correctness\u002Fscope — address before review; if observations — note and proceed |\n| **NEEDS_CONTEXT** | Provide the missing context and re-dispatch |\n| **BLOCKED** | Assess: context problem → re-dispatch with more context; too hard → escalate approach; too large → break down; plan wrong → escalate to human |\n\n**Never** ignore an escalation or force the same agent to retry without changes.\n\n## Red Flags\n\n**Never:**\n\n- Skip reviews (spec compliance OR code quality)\n- Proceed with unfixed issues\n- Dispatch multiple implementer agents in parallel (conflicts)\n- Make the Codex agent read the plan file (provide full text instead)\n- **Start code quality review before spec compliance passes** (wrong order)\n- Move to next task while either review has open issues\n\n**If reviewer finds issues:** resume the implementer thread to fix, reviewer reviews again, repeat until approved.\n\n## Integration\n\n- **superpowers-cc-to-codex:writing-plans-codex** — creates the plan this skill executes\n- **superpowers-cc-to-codex:requesting-code-review-codex** — code review via the `codex_review` MCP tool\n","\u003C!--\nUpstream source: obra\u002Fsuperpowers skills\u002Fsubagent-driven-development\u002FSKILL.md\nLast synced: 2026-04-03\nDivergence: Codex-backed implementer\u002Freviewer agents replace Claude subagents; adapter-managed resume flow\n-->\n\n# Subagent-Driven Development\n\nExecute plan by calling `codex_implement` per task via the MCP tool, with two-stage review after each: Claude-side spec compliance first, then `codex_review` for code quality.\n\n**Core principle:** Codex implementer per task + two-stage review (spec then quality) = high quality, fast iteration\n\n\u003CHARD-GATE>\nYou MUST dispatch codex_implement for every plan task (Step 3).\nYou cannot implement code yourself, even for simple tasks.\nSpec review requires seeing EXTERNAL implementation. No dispatch = no task completion.\n\u003C\u002FHARD-GATE>\n\n## Checklist\n\nYou MUST create a task for each plan task and complete them in order.\n\n**Setup:**\n\n1. **Read plan** — read the plan file once, extract all tasks with full text and context\n2. **Create TodoWrite** — one entry per plan task\n\n**Per task:**\n\n3. **Capture task base** — before dispatching the implementer, record the current revision:\n\n```bash\nTASK_BASE_SHA=\"$(git rev-parse HEAD)\"\n```\n\nSubstitute the captured TASK_BASE_SHA value as the structured review base for this task, so the reviewer inspects only the task's delta rather than all prior tasks.\n\n4. **Dispatch implementer** — call the `codex_implement` MCP tool.\n   Pass the task description as the `prompt`:\n\n```json\n{\n  \"tool\": \"codex_implement\",\n  \"arguments\": {\n    \"taskId\": \"task-17\",\n    \"prompt\": \"Implement Task 3 from docs\u002Fsuperpowers\u002Fplans\u002F2026-04-03-agent-forwarding.md. Keep the work scoped to agents\u002Fcodex-implementer.md and tests\u002Fprompt-contracts\u002Fexecution-workflows.test.mjs.\",\n    \"workspaceRoot\": \"\u002Fabsolute\u002Fpath\u002Fto\u002Fyour\u002Frepo\"\n  }\n}\n```\n\n5. **Handle status** — see Status Handling below\n6. **Spec compliance review** — Claude reads the actual code and verifies against spec using `spec-review-template.md`; if issues remain, resume the same implementer thread with `codex_resume`:\n\n```json\n{\n  \"tool\": \"codex_resume\",\n  \"arguments\": {\n    \"taskId\": \"task-17\",\n    \"sessionId\": \"019d4f82-58b8-72d3-9212-2e3d3fc69bcb\",\n    \"prompt\": \"Fix these issues:\\n- The agent still lacks tools: Bash.\\n- The prompt-file override is ignored in the TDD path.\",\n    \"workspaceRoot\": \"\u002Fabsolute\u002Fpath\u002Fto\u002Fyour\u002Frepo\"\n  }\n}\n```\n\n7. **Code quality review** — call `codex_review` only after spec compliance passes:\n\n```json\n{\n  \"tool\": \"codex_review\",\n  \"arguments\": {\n    \"taskId\": \"task-17-review\",\n    \"reviewStyle\": \"structured\",\n    \"scope\": { \"kind\": \"base\", \"base\": \"\u003Ccaptured TASK_BASE_SHA value>\" },\n    \"prompt\": \"Review the implementation of Task 4 from docs\u002Fsuperpowers\u002Fplans\u002F2026-04-03-agent-forwarding.md. Focus on regressions, incorrect routing, and missing test coverage.\",\n    \"workspaceRoot\": \"\u002Fabsolute\u002Fpath\u002Fto\u002Fyour\u002Frepo\"\n  }\n}\n```\n\n8. **Mark complete** — only after both gates pass\n\n**After all tasks:**\n\n9. **Final review** — call `codex_review` for the entire implementation with a new task ID and the final base SHA.\n10. **Finish branch** — wrap up the development branch\n\n## Status Handling\n\nHandle the implementer's reported status:\n\n| Status | Action |\n|---|---|\n| **DONE** | Proceed to spec compliance review |\n| **DONE_WITH_CONCERNS** | Read concerns; if correctness\u002Fscope — address before review; if observations — note and proceed |\n| **NEEDS_CONTEXT** | Provide the missing context and re-dispatch |\n| **BLOCKED** | Assess: context problem → re-dispatch with more context; too hard → escalate approach; too large → break down; plan wrong → escalate to human |\n\n**Never** ignore an escalation or force the same agent to retry without changes.\n\n## Red Flags\n\n**Never:**\n\n- Skip reviews (spec compliance OR code quality)\n- Proceed with unfixed issues\n- Dispatch multiple implementer agents in parallel (conflicts)\n- Make the Codex agent read the plan file (provide full text instead)\n- **Start code quality review before spec compliance passes** (wrong order)\n- Move to next task while either review has open issues\n\n**If reviewer finds issues:** resume the implementer thread to fix, reviewer reviews again, repeat until approved.\n\n## Integration\n\n- **superpowers-cc-to-codex:writing-plans-codex** — creates the plan this skill executes\n- **superpowers-cc-to-codex:requesting-code-review-codex** — code review via the `codex_review` MCP tool\n",{"name":5,"description":7,"disable-model-invocation":41},true,"main","687848074913d23ee93673a275fcf3e2715c6d57","skills\u002Fsubagent-driven-development-codex",2,"https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex","https:\u002F\u002Fgithub.com\u002Fmzored","2026-05-01 10:50:51",{"source":8,"skillName":5,"commands":50,"prompt":60,"zipAvailable":41},[51,54,57],{"tool":52,"command":53},"npx","npx skills add mzored\u002Fsuperpowers-cc-to-codex --skill subagent-driven-development-codex",{"tool":55,"command":56},"bunx","bunx skills add mzored\u002Fsuperpowers-cc-to-codex --skill subagent-driven-development-codex",{"tool":58,"command":59},"pnpm","pnpm dlx skills add mzored\u002Fsuperpowers-cc-to-codex --skill subagent-driven-development-codex","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Fskills\u002Fsubagent-driven-development-codex\u002FSKILL.md",[62,68,72],{"path":63,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":66,"sourceHtmlUrl":67},"prompts\u002Ffix-task.md","REFERENCE","参考文档",158,"https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Fskills\u002Fsubagent-driven-development-codex\u002Fprompts\u002Ffix-task.md",{"path":69,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":70,"sourceHtmlUrl":71},"prompts\u002Fimplement-task.md",3237,"https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Fskills\u002Fsubagent-driven-development-codex\u002Fprompts\u002Fimplement-task.md",{"path":73,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":74,"sourceHtmlUrl":75},"spec-review-template.md",1273,"https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Fskills\u002Fsubagent-driven-development-codex\u002Fspec-review-template.md",[77,99,105,121,127,144],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":93,"publishedAt":36},"01M6KKKN4DG8CRDF02C85KQQZW","writing-plans","Upstream fixture","https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fupstream-superpowers\u002Fskills\u002Fwriting-plans\u002FSKILL.md",[83,88],{"categoryNo":84,"categoryName":85,"taxonomyType":86,"taxonomyTypeDesc":87},"01JBSEEDCAT000000000000018","写作与文案","SCENARIO","使用场景",{"categoryNo":89,"categoryName":90,"taxonomyType":91,"taxonomyTypeDesc":92},"01JBSEEDCAT000000000000036","内容创作者","PROFESSION","适用职业",[94,95,96,97,98],"plans","writing plans","writing","upstream","fixture",{"skillNo":100,"name":101,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":36},"01M6KKKHN2H20G5B7GG39JPKNS","subagent-driven-development","https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fupstream-superpowers\u002Fskills\u002Fsubagent-driven-development\u002FSKILL.md",[],[21,22,23,24,97,98],{"skillNo":106,"name":107,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":116,"publishedAt":36},"01M6KKKDQJ9SEXY9KQ5CKESHJ2","requesting-code-review","https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fupstream-superpowers\u002Fskills\u002Frequesting-code-review\u002FSKILL.md",[110,113],{"categoryNo":111,"categoryName":112,"taxonomyType":86,"taxonomyTypeDesc":87},"01JBSEEDCAT000000000000007","代码审查",{"categoryNo":114,"categoryName":115,"taxonomyType":91,"taxonomyTypeDesc":92},"01JBSEEDCAT000000000000030","开发工程师",[117,118,119,120,97,98],"requesting code review","code","review","code review",{"skillNo":122,"name":123,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":124,"categories":125,"keywords":126,"publishedAt":36},"01M6KKK9G8DB9QVWP07F1NK5QA","brainstorming","https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fupstream-superpowers\u002Fskills\u002Fbrainstorming\u002FSKILL.md",[],[123,98,97],{"skillNo":128,"name":129,"nameZh":6,"description":130,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":131,"categories":132,"keywords":135,"publishedAt":36},"01M6KKK40C8NTEV2QT1RX1WWG6","writing-plans-codex","Use when you have a spec or requirements for a multi-step task, before touching code. Codex-backed variant.","https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Fskills\u002Fwriting-plans-codex\u002FSKILL.md",[133,134],{"categoryNo":84,"categoryName":85,"taxonomyType":86,"taxonomyTypeDesc":87},{"categoryNo":89,"categoryName":90,"taxonomyType":91,"taxonomyTypeDesc":92},[136,94,95,19,96,33,137,32,138,139,140,141,142,143,118,35],"writing plans codex","for","you","multi-step task","multi","requirements","task","use",{"skillNo":145,"name":146,"nameZh":6,"description":147,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":148,"categories":149,"keywords":150,"publishedAt":36},"01M6KKJYHHPRGMBZ6WZ23WZXYR","verification-before-completion-codex","Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always","https:\u002F\u002Fgithub.com\u002Fmzored\u002Fsuperpowers-cc-to-codex\u002Fblob\u002Fmain\u002Fskills\u002Fverification-before-completion-codex\u002FSKILL.md",[],[151,19,152,153,154,155,156,157,158,159,143,160,161,162],"verification","verification before completion","verification before completion codex","output","claims","complete","commands","claim","evidence","work","about","assertions"]