{"id":"4d1177b9-5ddd-4225-bb08-011f0e041a89","shortId":"ctQZkJ","kind":"skill","title":"subagent-driven-development","tagline":"Use when executing implementation plans with independent tasks in the current session","description":"# Subagent-Driven Development\n\nExecute plan by dispatching fresh subagent per task, with two-stage review after each: spec compliance review first, then code quality review.\n\n**Core principle:** Fresh subagent per task + two-stage review (spec then quality) = high quality, fast iteration\n\n## When to Use\n```dot\ndigraph when_to_use {\n    \"Have implementation plan?\" [shape=diamond];\n    \"Tasks mostly independent?\" [shape=diamond];\n    \"Stay in this session?\" [shape=diamond];\n    \"subagent-driven-development\" [shape=box];\n    \"executing-plans\" [shape=box];\n    \"Manual execution or brainstorm first\" [shape=box];\n\n    \"Have implementation plan?\" -> \"Tasks mostly independent?\" [label=\"yes\"];\n    \"Have implementation plan?\" -> \"Manual execution or brainstorm first\" [label=\"no\"];\n    \"Tasks mostly independent?\" -> \"Stay in this session?\" [label=\"yes\"];\n    \"Tasks mostly independent?\" -> \"Manual execution or brainstorm first\" [label=\"no - tightly coupled\"];\n    \"Stay in this session?\" -> \"subagent-driven-development\" [label=\"yes\"];\n    \"Stay in this session?\" -> \"executing-plans\" [label=\"no - parallel session\"];\n}\n```\n\n**vs. Executing Plans (parallel session):**\n- Same session (no context switch)\n- Fresh subagent per task (no context pollution)\n- Two-stage review after each task: spec compliance first, then code quality\n- Faster iteration (no human-in-loop between tasks)\n\n## The Process\n\n```dot\ndigraph process {\n    rankdir=TB;\n\n    subgraph cluster_per_task {\n        label=\"Per Task\";\n        \"Dispatch implementer subagent (./implementer-prompt.md)\" [shape=box];\n        \"Implementer subagent asks questions?\" [shape=diamond];\n        \"Answer questions, provide context\" [shape=box];\n        \"Implementer subagent implements, tests, commits, self-reviews\" [shape=box];\n        \"Dispatch spec reviewer subagent (./spec-reviewer-prompt.md)\" [shape=box];\n        \"Spec reviewer subagent confirms code matches spec?\" [shape=diamond];\n        \"Implementer subagent fixes spec gaps\" [shape=box];\n        \"Dispatch code quality reviewer subagent (./code-quality-reviewer-prompt.md)\" [shape=box];\n        \"Code quality reviewer subagent approves?\" [shape=diamond];\n        \"Implementer subagent fixes quality issues\" [shape=box];\n        \"Mark task complete in TodoWrite\" [shape=box];\n    }\n\n    \"Read plan, extract all tasks with full text, note context, create TodoWrite\" [shape=box];\n    \"More tasks remain?\" [shape=diamond];\n    \"Dispatch final code reviewer subagent for entire implementation\" [shape=box];\n    \"Use superpowers:finishing-a-development-branch\" [shape=box style=filled fillcolor=lightgreen];\n\n    \"Read plan, extract all tasks with full text, note context, create TodoWrite\" -> \"Dispatch implementer subagent (./implementer-prompt.md)\";\n    \"Dispatch implementer subagent (./implementer-prompt.md)\" -> \"Implementer subagent asks questions?\";\n    \"Implementer subagent asks questions?\" -> \"Answer questions, provide context\" [label=\"yes\"];\n    \"Answer questions, provide context\" -> \"Dispatch implementer subagent (./implementer-prompt.md)\";\n    \"Implementer subagent asks questions?\" -> \"Implementer subagent implements, tests, commits, self-reviews\" [label=\"no\"];\n    \"Implementer subagent implements, tests, commits, self-reviews\" -> \"Dispatch spec reviewer subagent (./spec-reviewer-prompt.md)\";\n    \"Dispatch spec reviewer subagent (./spec-reviewer-prompt.md)\" -> \"Spec reviewer subagent confirms code matches spec?\";\n    \"Spec reviewer subagent confirms code matches spec?\" -> \"Implementer subagent fixes spec gaps\" [label=\"no\"];\n    \"Implementer subagent fixes spec gaps\" -> \"Dispatch spec reviewer subagent (./spec-reviewer-prompt.md)\" [label=\"re-review\"];\n    \"Spec reviewer subagent confirms code matches spec?\" -> \"Dispatch code quality reviewer subagent (./code-quality-reviewer-prompt.md)\" [label=\"yes\"];\n    \"Dispatch code quality reviewer subagent (./code-quality-reviewer-prompt.md)\" -> \"Code quality reviewer subagent approves?\";\n    \"Code quality reviewer subagent approves?\" -> \"Implementer subagent fixes quality issues\" [label=\"no\"];\n    \"Implementer subagent fixes quality issues\" -> \"Dispatch code quality reviewer subagent (./code-quality-reviewer-prompt.md)\" [label=\"re-review\"];\n    \"Code quality reviewer subagent approves?\" -> \"Mark task complete in TodoWrite\" [label=\"yes\"];\n    \"Mark task complete in TodoWrite\" -> \"More tasks remain?\";\n    \"More tasks remain?\" -> \"Dispatch implementer subagent (./implementer-prompt.md)\" [label=\"yes\"];\n    \"More tasks remain?\" -> \"Dispatch final code reviewer subagent for entire implementation\" [label=\"no\"];\n    \"Dispatch final code reviewer subagent for entire implementation\" -> \"Use superpowers:finishing-a-development-branch\";\n}\n```\n\n## Prompt Templates\n\n- `./implementer-prompt.md` - Dispatch implementer subagent\n- `./spec-reviewer-prompt.md` - Dispatch spec compliance reviewer subagent\n- `./code-quality-reviewer-prompt.md` - Dispatch code quality reviewer subagent\n\n## Example Workflow\n\n```\nYou: I'm using Subagent-Driven Development to execute this plan.\n\n[Read plan file once: docs/plans/feature-plan.md]\n[Extract all 5 tasks with full text and context]\n[Create TodoWrite with all tasks]\n\nTask 1: Hook installation script\n\n[Get Task 1 text and context (already extracted)]\n[Dispatch implementation subagent with full task text + context]\n\nImplementer: \"Before I begin - should the hook be installed at user or system level?\"\n\nYou: \"User level (~/.config/superpowers/hooks/)\"\n\nImplementer: \"Got it. Implementing now...\"\n[Later] Implementer:\n  - Implemented install-hook command\n  - Added tests, 5/5 passing\n  - Self-review: Found I missed --force flag, added it\n  - Committed\n\n[Dispatch spec compliance reviewer]\nSpec reviewer: ✅ Spec compliant - all requirements met, nothing extra\n\n[Get git SHAs, dispatch code quality reviewer]\nCode reviewer: Strengths: Good test coverage, clean. Issues: None. Approved.\n\n[Mark Task 1 complete]\n\nTask 2: Recovery modes\n\n[Get Task 2 text and context (already extracted)]\n[Dispatch implementation subagent with full task text + context]\n\nImplementer: [No questions, proceeds]\nImplementer:\n  - Added verify/repair modes\n  - 8/8 tests passing\n  - Self-review: All good\n  - Committed\n\n[Dispatch spec compliance reviewer]\nSpec reviewer: ❌ Issues:\n  - Missing: Progress reporting (spec says \"report every 100 items\")\n  - Extra: Added --json flag (not requested)\n\n[Implementer fixes issues]\nImplementer: Removed --json flag, added progress reporting\n\n[Spec reviewer reviews again]\nSpec reviewer: ✅ Spec compliant now\n\n[Dispatch code quality reviewer]\nCode reviewer: Strengths: Solid. Issues (Important): Magic number (100)\n\n[Implementer fixes]\nImplementer: Extracted PROGRESS_INTERVAL constant\n\n[Code reviewer reviews again]\nCode reviewer: ✅ Approved\n\n[Mark Task 2 complete]\n\n...\n\n[After all tasks]\n[Dispatch final code-reviewer]\nFinal reviewer: All requirements met, ready to merge\n\nDone!\n```\n\n## Advantages\n\n**vs. Manual execution:**\n- Subagents follow TDD naturally\n- Fresh context per task (no confusion)\n- Parallel-safe (subagents don't interfere)\n- Subagent can ask questions (before AND during work)\n\n**vs. Executing Plans:**\n- Same session (no handoff)\n- Continuous progress (no waiting)\n- Review checkpoints automatic\n\n**Efficiency gains:**\n- No file reading overhead (controller provides full text)\n- Controller curates exactly what context is needed\n- Subagent gets complete information upfront\n- Questions surfaced before work begins (not after)\n\n**Quality gates:**\n- Self-review catches issues before handoff\n- Two-stage review: spec compliance, then code quality\n- Review loops ensure fixes actually work\n- Spec compliance prevents over/under-building\n- Code quality ensures implementation is well-built\n\n**Cost:**\n- More subagent invocations (implementer + 2 reviewers per task)\n- Controller does more prep work (extracting all tasks upfront)\n- Review loops add iterations\n- But catches issues early (cheaper than debugging later)\n\n## Red Flags\n\n**Never:**\n- Skip reviews (spec compliance OR code quality)\n- Proceed with unfixed issues\n- Dispatch multiple implementation subagents in parallel (conflicts)\n- Make subagent read plan file (provide full text instead)\n- Skip scene-setting context (subagent needs to understand where task fits)\n- Ignore subagent questions (answer before letting them proceed)\n- Accept \"close enough\" on spec compliance (spec reviewer found issues = not done)\n- Skip review loops (reviewer found issues = implementer fixes = review again)\n- Let implementer self-review replace actual review (both are needed)\n- **Start code quality review before spec compliance is ✅** (wrong order)\n- Move to next task while either review has open issues\n\n**If subagent asks questions:**\n- Answer clearly and completely\n- Provide additional context if needed\n- Don't rush them into implementation\n\n**If reviewer finds issues:**\n- Implementer (same subagent) fixes them\n- Reviewer reviews again\n- Repeat until approved\n- Don't skip the re-review\n\n**If subagent fails task:**\n- Dispatch fix subagent with specific instructions\n- Don't try to fix manually (context pollution)\n\n## Integration\n\n**Required workflow skills:**\n- **superpowers:writing-plans** - Creates the plan this skill executes\n- **superpowers:requesting-code-review** - Code review template for reviewer subagents\n- **superpowers:finishing-a-development-branch** - Complete development after all tasks\n\n**Subagents should use:**\n- **superpowers:test-driven-development** - Subagents follow TDD for each task\n\n**Alternative workflow:**\n- **superpowers:executing-plans** - Use for parallel session instead of same-session execution\n\n## Limitations\n- Use this skill only when the task clearly matches the scope described above.\n- Do not treat the output as a substitute for environment-specific validation, testing, or expert review.\n- Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.","tags":["subagent","driven","development","antigravity","awesome","skills","sickn33","agent-skills","agentic-skills","ai-agent-skills","ai-agents","ai-coding"],"capabilities":["skill","source-sickn33","skill-subagent-driven-development","topic-agent-skills","topic-agentic-skills","topic-ai-agent-skills","topic-ai-agents","topic-ai-coding","topic-ai-workflows","topic-antigravity","topic-antigravity-skills","topic-claude-code","topic-claude-code-skills","topic-codex-cli","topic-codex-skills"],"categories":["antigravity-awesome-skills"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/sickn33/antigravity-awesome-skills/subagent-driven-development","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add sickn33/antigravity-awesome-skills","source_repo":"https://github.com/sickn33/antigravity-awesome-skills","install_from":"skills.sh"}},"qualityScore":"0.700","qualityRationale":"deterministic score 0.70 from registry signals: · indexed on github topic:agent-skills · 34460 github stars · SKILL.md body (9,974 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-04-22T06:51:56.758Z","embedding":null,"createdAt":"2026-04-18T21:45:37.055Z","updatedAt":"2026-04-22T06:51:56.758Z","lastSeenAt":"2026-04-22T06:51:56.758Z","tsv":"'/.config/superpowers/hooks':646 '/code-quality-reviewer-prompt.md':272,459,467,495,569 '/implementer-prompt.md':219,353,357,379,526,559 '/spec-reviewer-prompt.md':248,406,411,442,563 '1':609,615,706 '100':759,798 '2':709,714,815,947 '5':596 '5/5':661 '8/8':736 'accept':1022 'actual':928,1050 'ad':659,671,733,762,774 'add':962 'addit':1084 'advantag':834 'alreadi':619,718 'altern':1184 'answer':228,366,372,1017,1079 'approv':279,472,477,504,703,812,1108 'ask':224,360,364,382,857,1077,1233 'automat':876 'begin':632,903 'boundari':1241 'box':90,95,102,221,233,243,250,266,274,288,295,309,324,333 'brainstorm':99,117,136 'branch':331,556,1164 'built':941 'catch':911,965 'cheaper':968 'checkpoint':875 'clarif':1235 'clean':700 'clear':1080,1208 'close':1023 'cluster':210 'code':41,191,255,268,275,317,416,423,451,455,463,468,473,491,500,534,544,571,691,694,787,790,806,810,823,922,934,980,1056,1151,1153 'code-review':822 'command':658 'commit':238,388,398,673,744 'complet':291,507,514,707,816,896,1082,1165 'complianc':37,188,566,676,747,920,931,978,1027,1061 'compliant':681,784 'confirm':254,415,422,450 'conflict':992 'confus':847 'constant':805 'context':171,178,231,305,347,369,375,602,618,628,717,727,843,891,1006,1085,1132 'continu':870 'control':883,887,951 'core':44 'cost':942 'coupl':141 'coverag':699 'creat':306,348,603,1142 'criteria':1244 'curat':888 'current':15 'debug':970 'describ':1212 'develop':4,20,88,149,330,555,584,1163,1166,1177 'diamond':73,78,84,227,259,281,314 'digraph':65,205 'dispatch':24,216,244,267,315,350,354,376,402,407,438,454,462,490,523,532,542,560,564,570,621,674,690,720,745,786,820,986,1120 'docs/plans/feature-plan.md':593 'done':833,1033 'dot':64,204 'driven':3,19,87,148,583,1176 'earli':967 'effici':877 'either':1070 'enough':1024 'ensur':926,936 'entir':321,538,548 'environ':1224 'environment-specif':1223 'everi':758 'exact':889 'exampl':575 'execut':7,21,92,97,115,134,157,164,586,837,864,1147,1188,1199 'executing-plan':91,156,1187 'expert':1229 'extra':686,761 'extract':298,340,594,620,719,802,956 'fail':1118 'fast':59 'faster':193 'file':591,880,997 'fill':335 'fillcolor':336 'final':316,533,543,821,825 'find':1096 'finish':328,553,1161 'finishing-a-development-branch':327,552,1160 'first':39,100,118,137,189 'fit':1013 'fix':262,284,428,435,480,487,768,800,927,1041,1101,1121,1130 'flag':670,764,773,973 'follow':839,1179 'forc':669 'found':666,1030,1038 'fresh':25,46,173,842 'full':302,344,599,625,724,885,999 'gain':878 'gap':264,430,437 'gate':907 'get':613,687,712,895 'git':688 'good':697,743 'got':648 'handoff':869,914 'high':57 'hook':610,635,657 'human':197 'human-in-loop':196 'ignor':1014 'implement':8,70,104,112,217,222,234,236,260,282,322,351,355,358,362,377,380,384,386,394,396,426,433,478,485,524,539,549,561,622,629,647,650,653,654,721,728,732,767,770,799,801,937,946,988,1040,1045,1093,1098 'import':795 'independ':11,76,108,123,132 'inform':897 'input':1238 'instal':611,637,656 'install-hook':655 'instead':1001,1194 'instruct':1125 'integr':1134 'interfer':854 'interv':804 'invoc':945 'issu':286,482,489,701,751,769,794,912,966,985,1031,1039,1074,1097 'item':760 'iter':60,194,963 'json':763,772 'label':109,119,128,138,150,159,213,370,392,431,443,460,483,496,510,527,540 'later':652,971 'let':1019,1044 'level':642,645 'lightgreen':337 'limit':1200 'loop':199,925,961,1036 'm':579 'magic':796 'make':993 'manual':96,114,133,836,1131 'mark':289,505,512,704,813 'match':256,417,424,452,1209 'merg':832 'met':684,829 'miss':668,752,1246 'mode':711,735 'most':75,107,122,131 'move':1065 'multipl':987 'natur':841 'need':893,1008,1054,1087 'never':974 'next':1067 'none':702 'note':304,346 'noth':685 'number':797 'open':1073 'order':1064 'output':1218 'over/under-building':933 'overhead':882 'parallel':161,166,849,991,1192 'parallel-saf':848 'pass':662,738 'per':27,48,175,211,214,844,949 'permiss':1239 'plan':9,22,71,93,105,113,158,165,297,339,588,590,865,996,1141,1144,1189 'pollut':179,1133 'prep':954 'prevent':932 'principl':45 'proceed':731,982,1021 'process':203,206 'progress':753,775,803,871 'prompt':557 'provid':230,368,374,884,998,1083 'qualiti':42,56,58,192,269,276,285,456,464,469,474,481,488,492,501,572,692,788,906,923,935,981,1057 'question':225,229,361,365,367,373,383,730,858,899,1016,1078 'rankdir':207 're':445,498,1114 're-review':444,497,1113 'read':296,338,589,881,995 'readi':830 'recoveri':710 'red':972 'remain':312,519,522,531 'remov':771 'repeat':1106 'replac':1049 'report':754,757,776 'request':766,1150 'requesting-code-review':1149 'requir':683,828,1135,1237 'review':33,38,43,53,183,241,246,252,270,277,318,391,401,404,409,413,420,440,446,448,457,465,470,475,493,499,502,535,545,567,573,665,677,679,693,695,741,748,750,778,779,782,789,791,807,808,811,824,826,874,910,918,924,948,960,976,1029,1035,1037,1042,1048,1051,1058,1071,1095,1103,1104,1115,1152,1154,1157,1230 'rush':1090 'safe':850 'safeti':1240 'same-sess':1196 'say':756 'scene':1004 'scene-set':1003 'scope':1211 'script':612 'self':240,390,400,664,740,909,1047 'self-review':239,389,399,663,739,908,1046 'session':16,82,127,145,155,162,167,169,867,1193,1198 'set':1005 'shape':72,77,83,89,94,101,220,226,232,242,249,258,265,273,280,287,294,308,313,323,332 'shas':689 'skill':1137,1146,1203 'skill-subagent-driven-development' 'skip':975,1002,1034,1111 'solid':793 'source-sickn33' 'spec':36,54,187,245,251,257,263,403,408,412,418,419,425,429,436,439,447,453,565,675,678,680,746,749,755,777,781,783,919,930,977,1026,1028,1060 'specif':1124,1225 'stage':32,52,182,917 'start':1055 'stay':79,124,142,152 'stop':1231 'strength':696,792 'style':334 'subag':2,18,26,47,86,147,174,218,223,235,247,253,261,271,278,283,319,352,356,359,363,378,381,385,395,405,410,414,421,427,434,441,449,458,466,471,476,479,486,494,503,525,536,546,562,568,574,582,623,722,838,851,855,894,944,989,994,1007,1015,1076,1100,1117,1122,1158,1170,1178 'subagent-driven':17,581 'subagent-driven-develop':1,85,146 'subgraph':209 'substitut':1221 'success':1243 'superpow':326,551,1138,1148,1159,1173,1186 'surfac':900 'switch':172 'system':641 'task':12,28,49,74,106,121,130,176,186,201,212,215,290,300,311,342,506,513,518,521,530,597,607,608,614,626,705,708,713,725,814,819,845,950,958,1012,1068,1119,1169,1183,1207 'tb':208 'tdd':840,1180 'templat':558,1155 'test':237,387,397,660,698,737,1175,1227 'test-driven-develop':1174 'text':303,345,600,616,627,715,726,886,1000 'tight':140 'todowrit':293,307,349,509,516,604 'topic-agent-skills' 'topic-agentic-skills' 'topic-ai-agent-skills' 'topic-ai-agents' 'topic-ai-coding' 'topic-ai-workflows' 'topic-antigravity' 'topic-antigravity-skills' 'topic-claude-code' 'topic-claude-code-skills' 'topic-codex-cli' 'topic-codex-skills' 'treat':1216 'tri':1128 'two':31,51,181,916 'two-stag':30,50,180,915 'understand':1010 'unfix':984 'upfront':898,959 'use':5,63,68,325,550,580,1172,1190,1201 'user':639,644 'valid':1226 'verify/repair':734 'vs':163,835,863 'wait':873 'well':940 'well-built':939 'work':862,902,929,955 'workflow':576,1136,1185 'write':1140 'writing-plan':1139 'wrong':1063 'yes':110,129,151,371,461,511,528","prices":[{"id":"7c7a0b53-fd80-4be6-89e3-6de244725e6c","listingId":"4d1177b9-5ddd-4225-bb08-011f0e041a89","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"sickn33","category":"antigravity-awesome-skills","install_from":"skills.sh"},"createdAt":"2026-04-18T21:45:37.055Z"}],"sources":[{"listingId":"4d1177b9-5ddd-4225-bb08-011f0e041a89","source":"github","sourceId":"sickn33/antigravity-awesome-skills/subagent-driven-development","sourceUrl":"https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/subagent-driven-development","isPrimary":false,"firstSeenAt":"2026-04-18T21:45:37.055Z","lastSeenAt":"2026-04-22T06:51:56.758Z"}],"details":{"listingId":"4d1177b9-5ddd-4225-bb08-011f0e041a89","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"sickn33","slug":"subagent-driven-development","github":{"repo":"sickn33/antigravity-awesome-skills","stars":34460,"topics":["agent-skills","agentic-skills","ai-agent-skills","ai-agents","ai-coding","ai-workflows","antigravity","antigravity-skills","claude-code","claude-code-skills","codex-cli","codex-skills","cursor","cursor-skills","developer-tools","gemini-cli","gemini-skills","kiro","mcp","skill-library"],"license":"mit","html_url":"https://github.com/sickn33/antigravity-awesome-skills","pushed_at":"2026-04-22T06:40:00Z","description":"Installable GitHub library of 1,400+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and more. Includes installer CLI, bundles, workflows, and official/community skill collections.","skill_md_sha":"a0238e025d7097c0f97b1209214c70858f83d795","skill_md_path":"skills/subagent-driven-development/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/subagent-driven-development"},"layout":"multi","source":"github","category":"antigravity-awesome-skills","frontmatter":{"name":"subagent-driven-development","description":"Use when executing implementation plans with independent tasks in the current session"},"skills_sh_url":"https://skills.sh/sickn33/antigravity-awesome-skills/subagent-driven-development"},"updatedAt":"2026-04-22T06:51:56.758Z"}}