{"id":"59e1a3cf-b270-4cf3-b976-b7b322122e33","shortId":"BaDXUT","kind":"skill","title":"recipe-reverse-engineer","tagline":"Generate PRD and Design Docs from existing codebase through discovery, generation, verification, and review workflow","description":"**Context**: Reverse engineering workflow to create documentation from existing code\n\nTarget: $ARGUMENTS\n\n## Orchestrator Definition\n\n**Core Identity**: \"I am an orchestrator.\"\n\n**Execution Protocol**:\n1. **Delegate all work through Agent tool** — invoke sub-agents, pass deliverable paths between them, and report results (permitted tools: see subagents-orchestration-guide \"Orchestrator's Permitted Tools\")\n2. **Process one step at a time**: Execute steps sequentially within each unit (2 → 3 → 4 → 5). Each step's output is the required input for the next step. Complete all steps for one unit before starting the next\n3. **Pass `$STEP_N_OUTPUT` as-is** to sub-agents — the orchestrator bridges data without processing or filtering it\n\n**Task Registration**: Register phases first using TaskCreate, then steps within each phase as you enter it. Update status using TaskUpdate.\n\n## Step 0: Initial Configuration\n\n### 0.1 Scope Confirmation\n\nUse AskUserQuestion to confirm:\n1. **Target path**: Which directory/module to document\n2. **Depth**: PRD only, or PRD + Design Docs\n3. **Reference Architecture**: layered / mvc / clean / hexagonal / none\n4. **Human review**: Yes (recommended) / No (fully autonomous)\n5. **Fullstack design**: Yes / No\n   - Yes: For each functional unit, generate backend + frontend Design Docs\n   - Note: Requires both agents (technical-designer, technical-designer-frontend)\n\n### 0.2 Output Configuration\n\n- PRD output: `docs/prd/` or existing PRD directory\n- Design Doc output: `docs/design/` or existing design directory\n- Verify directories exist, create if needed\n\n## Workflow Overview\n\n```\nPhase 1: PRD Generation\n  Step 1: Scope Discovery (unified, single pass → group into PRD units → human review)\n  Step 2-5: Per-unit loop (Generation → Verification → Review → Revision)\n\nPhase 2: Design Doc Generation (if requested)\n  Step 6: Design Doc Scope Mapping (reuse Step 1 results, no re-discovery)\n  Step 7-10: Per-unit loop (Generation → Verification → Review → Revision)\n  ※ fullstack=Yes: each unit produces backend + frontend Design Docs\n```\n\n## Phase 1: PRD Generation\n\n**Register using TaskCreate**:\n- Step 1: PRD Scope Discovery\n- Per-unit processing (Steps 2-5 for each unit)\n\n### Step 1: PRD Scope Discovery\n\n**Agent tool invocation**:\n```\nsubagent_type: dev-workflows:scope-discoverer\ndescription: \"Discover functional scope\"\nprompt: |\n  Discover functional scope targets in the codebase.\n\n  target_path: $USER_TARGET_PATH\n  reference_architecture: $USER_RA_CHOICE\n  focus_area: $USER_FOCUS_AREA (if specified)\n```\n\n**Store output as**: `$STEP_1_OUTPUT`\n\n**Quality Gate**:\n- At least one unit discovered → proceed\n- No units discovered → ask user for hints\n- `$STEP_1_OUTPUT.prdUnits` exists\n- All `sourceUnits` across `prdUnits` (flattened, deduplicated) match the set of `discoveredUnits` IDs — no unit missing, no unit duplicated\n- Each discovered unit's `unitInventory` has at least one non-empty category (routes, testFiles, or publicExports). Units with all three empty indicate incomplete discovery — re-run scope-discoverer with focus on that unit's relatedFiles\n\n**Human Review Point** (if enabled): Present `$STEP_1_OUTPUT.prdUnits` with their source unit mapping. The user confirms, adjusts grouping, or excludes units from scope. This is the most important review point — incorrect grouping cascades into all downstream documents.\n\n### Step 2-5: Per-Unit Processing\n\n**FOR** each unit in `$STEP_1_OUTPUT.prdUnits` **(sequential, one unit at a time)**:\n\n#### Step 2: PRD Generation\n\n**Agent tool invocation**:\n```\nsubagent_type: dev-workflows:prd-creator\ndescription: \"Generate PRD\"\nprompt: |\n  Create reverse-engineered PRD for the following feature.\n\n  Operation Mode: reverse-engineer\n  External Scope Provided: true\n\n  Feature: $PRD_UNIT_NAME (from $STEP_1_OUTPUT)\n  Description: $PRD_UNIT_DESCRIPTION\n  Related Files: $PRD_UNIT_COMBINED_RELATED_FILES\n  Entry Points: $PRD_UNIT_COMBINED_ENTRY_POINTS\n\n  Use provided scope as investigation starting point.\n  If tracing entry points reveals files outside this scope, include them.\n  Create final version PRD based on thorough code investigation.\n```\n\n**Store output as**: `$STEP_2_OUTPUT` (PRD path)\n\n#### Step 3: Code Verification\n\n**Prerequisite**: $STEP_2_OUTPUT (PRD path from Step 2)\n\n**Agent tool invocation**:\n```\nsubagent_type: dev-workflows:code-verifier\ndescription: \"Verify PRD consistency\"\nprompt: |\n  Verify consistency between PRD and code implementation.\n\n  doc_type: prd\n  document_path: $STEP_2_OUTPUT\n  verbose: false\n```\n\nNote: Omit `code_paths` — the verifier independently discovers code scope from the document, ensuring independent verification not constrained by scope-discoverer's output.\n\n**Store output as**: `$STEP_3_OUTPUT`\n\n**Quality Gate**:\n- consistencyScore >= 70 AND verifiableClaimCount >= 20 → proceed to review\n- consistencyScore >= 70 BUT verifiableClaimCount < 20 → re-run verifier (investigation too shallow)\n- consistencyScore < 70 → flag for detailed review\n\n#### Step 4: Review\n\n**Required Input**: $STEP_3_OUTPUT (verification JSON from Step 3)\n\n**Agent tool invocation**:\n```\nsubagent_type: dev-workflows:document-reviewer\ndescription: \"Review PRD\"\nprompt: |\n  Review the following PRD considering code verification findings.\n\n  doc_type: PRD\n  target: $STEP_2_OUTPUT\n  mode: composite\n  code_verification: $STEP_3_OUTPUT\n\n  ## Additional Review Focus\n  - Alignment between PRD claims and verification evidence\n  - Resolution recommendations for each discrepancy\n  - Completeness of undocumented feature coverage\n```\n\n**Store output as**: `$STEP_4_OUTPUT`\n\n#### Step 5: Revision (conditional)\n\n**Trigger Conditions** (any one of the following):\n- Review status is \"Needs Revision\" or \"Rejected\"\n- Critical discrepancies exist in `$STEP_3_OUTPUT`\n- consistencyScore < 70\n\n**Agent tool invocation**:\n```\nsubagent_type: dev-workflows:prd-creator\ndescription: \"Revise PRD\"\nprompt: |\n  Update PRD based on review feedback and code verification results.\n\n  Operation Mode: update\n  Existing PRD: $STEP_2_OUTPUT\n\n  ## Review Feedback\n  $STEP_4_OUTPUT\n\n  ## Code Verification Results\n  $STEP_3_OUTPUT\n\n  Address discrepancies by severity. Critical and major items require correction.\n  Minor items: correct if straightforward, otherwise leave as-is with rationale.\n```\n\n**Loop Control**: Maximum 2 revision cycles. After 2 cycles, flag for human review regardless of status.\n\n#### Unit Completion\n\n- [ ] Review status is \"Approved\" or \"Approved with Conditions\"\n- [ ] Human review passed (if enabled in Step 0)\n\n**Next**: Proceed to next unit. After all units → Phase 2.\n\n## Phase 2: Design Doc Generation\n\n*Execute only if Design Docs were requested in Step 0*\n\n**Register using TaskCreate**:\n- Step 6: Design Doc Scope Mapping\n- Per-unit processing (Steps 7-10 for each unit)\n\n### Step 6: Design Doc Scope Mapping\n\n**No additional discovery required.** Use `$STEP_1_OUTPUT.discoveredUnits` (implementation-granularity units) for technical profiles. Use `$STEP_1_OUTPUT.prdUnits[].sourceUnits` to trace which discovered units belong to each PRD unit.\n\nEach PRD unit from Phase 1 maps to Design Doc unit(s):\n- **Standard mode (fullstack=No)**: 1 PRD unit → 1 Design Doc (using technical-designer)\n- **Fullstack mode (fullstack=Yes)**: 1 PRD unit → 2 Design Docs (technical-designer + technical-designer-frontend)\n\nMap `$STEP_1_OUTPUT` units to Design Doc generation targets, carrying forward:\n- `technicalProfile.primaryModules` → Primary Files\n- `technicalProfile.publicInterfaces` → Public Interfaces\n- `dependencies` → Dependencies\n- `relatedFiles` → Scope boundary\n- `unitInventory` → Unit Inventory (routes, test files, public exports)\n\n**Store output as**: `$STEP_6_OUTPUT`\n\n### Step 7-10: Per-Unit Processing\n\n**FOR** each unit in `$STEP_6_OUTPUT` **(sequential, one unit at a time)**:\n\n#### Step 7: Design Doc Generation\n\n**Scope**: Document the current architecture exactly as implemented in code.\n\n**Standard mode (fullstack=No)**:\n\n**Agent tool invocation**:\n```\nsubagent_type: dev-workflows:technical-designer\ndescription: \"Generate Design Doc\"\nprompt: |\n  Create Design Doc for the following feature based on existing code.\n\n  Operation Mode: reverse-engineer\n\n  Feature: $UNIT_NAME (from $STEP_6_OUTPUT)\n  Description: $UNIT_DESCRIPTION\n  Primary Files: $UNIT_PRIMARY_MODULES\n  Public Interfaces: $UNIT_PUBLIC_INTERFACES\n  Dependencies: $UNIT_DEPENDENCIES\n  Unit Inventory: $UNIT_INVENTORY (routes, test files, public exports from scope discovery)\n\n  Parent PRD: $APPROVED_PRD_PATH\n\n  Document current architecture as-is. Use Unit Inventory as a completeness baseline — all routes and exports should be accounted for in the Design Doc.\n```\n\n**Store output as**: `$STEP_7_OUTPUT`\n\n**Fullstack mode (fullstack=Yes)**:\n\nFor each unit, invoke 7a then 7b sequentially (7b depends on 7a output):\n\n**7a. Backend Design Doc**:\n```\nsubagent_type: dev-workflows:technical-designer\ndescription: \"Generate backend Design Doc\"\nprompt: |\n  Create a backend Design Doc for the following feature based on existing code.\n\n  Operation Mode: reverse-engineer\n\n  Feature: $UNIT_NAME (from $STEP_6_OUTPUT)\n  Description: $UNIT_DESCRIPTION\n  Primary Files: $UNIT_PRIMARY_MODULES\n  Public Interfaces: $UNIT_PUBLIC_INTERFACES\n  Dependencies: $UNIT_DEPENDENCIES\n  Unit Inventory: $UNIT_INVENTORY\n\n  Parent PRD: $APPROVED_PRD_PATH\n\n  Focus on: API contracts, data layer, business logic, service architecture.\n  Document current architecture as-is. Use Unit Inventory as completeness baseline.\n```\n\n**Store output as**: `$STEP_7a_OUTPUT`\n\n**7b. Frontend Design Doc**:\n```\nsubagent_type: dev-workflows-frontend:technical-designer-frontend\ndescription: \"Generate frontend Design Doc\"\nprompt: |\n  Create a frontend Design Doc for the following feature based on existing code.\n\n  Operation Mode: reverse-engineer\n\n  Feature: $UNIT_NAME (from $STEP_6_OUTPUT)\n  Description: $UNIT_DESCRIPTION\n  Primary Files: $UNIT_PRIMARY_MODULES\n  Public Interfaces: $UNIT_PUBLIC_INTERFACES\n  Dependencies: $UNIT_DEPENDENCIES\n  Unit Inventory: $UNIT_INVENTORY\n\n  Parent PRD: $APPROVED_PRD_PATH\n  Backend Design Doc: $STEP_7a_OUTPUT\n\n  Reference backend Design Doc for API contracts.\n  Focus on: component hierarchy, state management, UI interactions, data fetching.\n  Document current architecture as-is. Use Unit Inventory as completeness baseline.\n```\n\n**Store output as**: `$STEP_7b_OUTPUT`\n\n#### Step 8: Code Verification\n\n**Standard mode**: Verify `$STEP_7_OUTPUT`.\n\n**Fullstack mode**: Verify each Design Doc separately.\n\n**Agent tool invocation (per Design Doc)**:\n```\nsubagent_type: dev-workflows:code-verifier\ndescription: \"Verify Design Doc consistency\"\nprompt: |\n  Verify consistency between Design Doc and code implementation.\n\n  doc_type: design-doc\n  document_path: $STEP_7_OUTPUT (or $STEP_7a_OUTPUT / $STEP_7b_OUTPUT)\n  verbose: false\n```\n\nNote: Omit `code_paths` — the verifier independently discovers code scope from the document.\n\n**Store output as**: `$STEP_8_OUTPUT`\n\n#### Step 9: Review\n\n**Required Input**: $STEP_8_OUTPUT (verification JSON from Step 8)\n\n**Agent tool invocation (per Design Doc)**:\n```\nsubagent_type: dev-workflows:document-reviewer\ndescription: \"Review Design Doc\"\nprompt: |\n  Review the following Design Doc considering code verification findings.\n\n  doc_type: DesignDoc\n  target: $STEP_7_OUTPUT (or $STEP_7a_OUTPUT / $STEP_7b_OUTPUT)\n  mode: composite\n  code_verification: $STEP_8_OUTPUT\n\n  ## Parent PRD\n  $APPROVED_PRD_PATH\n\n  ## Additional Review Focus\n  - Technical accuracy of documented interfaces\n  - Consistency with parent PRD scope\n  - Completeness of unit boundary definitions\n```\n\n**Store output as**: `$STEP_9_OUTPUT`\n\n#### Step 10: Revision (conditional)\n\n**Trigger Conditions** (same as Step 5):\n- Review status is \"Needs Revision\" or \"Rejected\"\n- Critical discrepancies exist in `$STEP_8_OUTPUT`\n- consistencyScore < 70\n\n**Agent tool invocation (per Design Doc)**:\n```\nsubagent_type: dev-workflows:technical-designer (or dev-workflows-frontend:technical-designer-frontend for frontend Design Docs)\ndescription: \"Revise Design Doc\"\nprompt: |\n  Update Design Doc based on review feedback and code verification results.\n\n  Operation Mode: update\n  Existing Design Doc: $STEP_7_OUTPUT (or $STEP_7a_OUTPUT / $STEP_7b_OUTPUT)\n\n  ## Review Feedback\n  $STEP_9_OUTPUT\n\n  ## Code Verification Results\n  $STEP_8_OUTPUT\n\n  Address discrepancies by severity. Critical and major items require correction.\n  Minor items: correct if straightforward, otherwise leave as-is with rationale.\n```\n\n**Loop Control**: Maximum 2 revision cycles. After 2 cycles, flag for human review regardless of status.\n\n#### Unit Completion\n\n- [ ] Review status is \"Approved\" or \"Approved with Conditions\"\n- [ ] Human review passed (if enabled in Step 0)\n\n**Next**: Proceed to next unit. After all units → Final Report.\n\n## Final Report\n\nOutput summary including:\n- Generated documents table (Type, Name, Consistency Score, Review Status)\n- Action items (critical discrepancies, undocumented features, flagged items)\n- Next steps checklist\n\n## Error Handling\n\n| Error | Action |\n|-------|--------|\n| Discovery finds nothing | Ask user for project structure hints |\n| Generation fails | Log failure, continue with other units, report in summary |\n| consistencyScore < 50 | Flag for mandatory human review — require explicit human approval |\n| Review rejects after 2 revisions | Stop loop, flag for human intervention |","tags":["recipe","reverse","engineer","claude","code","workflows","shinpr","agent-skills","agentic-ai","ai-agents","automation","claude-code"],"capabilities":["skill","source-shinpr","skill-recipe-reverse-engineer","topic-agent-skills","topic-agentic-ai","topic-ai-agents","topic-automation","topic-claude-code","topic-claude-code-plugin","topic-code-quality","topic-developer-tools","topic-development-workflow","topic-llm-orchestration","topic-productivity","topic-prompt-engineering"],"categories":["claude-code-workflows"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/shinpr/claude-code-workflows/recipe-reverse-engineer","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add shinpr/claude-code-workflows","source_repo":"https://github.com/shinpr/claude-code-workflows","install_from":"skills.sh"}},"qualityScore":"0.613","qualityRationale":"deterministic score 0.61 from registry signals: · indexed on github topic:agent-skills · 327 github stars · SKILL.md body (13,447 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":"2026-05-02T18:53:52.966Z","embedding":null,"createdAt":"2026-04-18T22:03:15.118Z","updatedAt":"2026-05-02T18:53:52.966Z","lastSeenAt":"2026-05-02T18:53:52.966Z","tsv":"'-10':297,960,1078 '-5':265,333,499 '0':153,919,944,1767 '0.1':156 '0.2':220 '1':42,163,247,251,289,316,323,338,386,558,1001,1012,1015,1026,1041 '10':1617 '2':72,85,170,264,275,332,498,516,609,619,625,655,758,851,889,893,929,931,1029,1737,1741,1841 '20':695,703 '3':86,111,178,614,687,723,729,765,816,862 '4':87,186,718,791,856 '5':88,194,794,1625 '50':1828 '6':282,949,965,1074,1088,1152,1276,1374 '7':296,959,1077,1097,1216,1450,1495,1571,1692 '70':692,700,712,819,1641 '7a':1226,1233,1235,1329,1405,1499,1575,1696 '7b':1228,1230,1331,1440,1502,1578,1699 '8':1443,1523,1531,1537,1585,1638,1710 '9':1526,1614,1704 'account':1206 'accuraci':1596 'across':407 'action':1792,1806 'addit':767,971,1592 'address':864,1712 'adjust':476 'agent':47,52,122,212,342,519,626,730,820,1115,1459,1538,1642 'align':770 'api':1305,1412 'approv':907,909,1184,1300,1398,1589,1755,1757,1837 'architectur':180,371,1105,1189,1312,1315,1426 'area':376,379 'argument':31 'as-i':116,881,1190,1316,1427,1729 'ask':399,1810 'askuserquest':160 'autonom':193 'backend':205,311,1236,1249,1255,1401,1408 'base':600,837,1138,1262,1360,1677 'baselin':1199,1324,1435 'belong':991 'boundari':1061,1608 'bridg':125 'busi':1309 'carri':1049 'cascad':492 'categori':435 'checklist':1802 'choic':374 'claim':773 'clean':183 'code':29,603,615,635,647,661,667,750,762,842,858,1110,1141,1265,1363,1444,1471,1485,1508,1514,1563,1582,1682,1706 'code-verifi':634,1470 'codebas':12,364 'combin':568,575 'complet':101,782,903,1198,1323,1434,1605,1751 'compon':1416 'composit':761,1581 'condit':796,798,911,1619,1621,1759 'configur':155,222 'confirm':158,162,475 'consid':749,1562 'consist':640,643,1477,1480,1600,1788 'consistencyscor':691,699,711,818,1640,1827 'constrain':676 'context':20 'continu':1820 'contract':1306,1413 'control':887,1735 'core':34 'correct':873,876,1721,1724 'coverag':786 'creat':25,241,534,596,1131,1253,1351 'creator':529,830 'critic':811,868,1633,1716,1794 'current':1104,1188,1314,1425 'cycl':891,894,1739,1742 'data':126,1307,1422 'dedupl':410 'definit':33,1609 'deleg':43 'deliver':54 'depend':1057,1058,1167,1169,1231,1291,1293,1389,1391 'depth':171 'descript':353,530,560,563,637,741,831,1126,1154,1156,1247,1278,1280,1345,1376,1378,1473,1552,1669 'design':8,176,196,207,215,218,230,236,276,283,313,932,938,950,966,1004,1016,1021,1030,1034,1037,1045,1098,1125,1128,1132,1210,1237,1246,1250,1256,1333,1343,1348,1354,1402,1409,1456,1463,1475,1482,1490,1542,1554,1560,1646,1655,1663,1667,1671,1675,1689 'design-doc':1489 'designdoc':1568 'detail':715 'dev':348,525,632,736,826,1121,1242,1338,1468,1547,1651,1658 'dev-workflow':347,524,631,735,825,1120,1241,1467,1546,1650 'dev-workflows-frontend':1337,1657 'directori':229,237,239 'directory/module':167 'discov':354,358,394,398,424,666,989,1513 'discover':352,453,680 'discoveredunit':415 'discoveri':14,253,294,326,341,447,972,1181,1807 'discrep':781,812,865,1634,1713,1795 'doc':9,177,208,231,277,284,314,649,753,933,939,951,967,1005,1017,1031,1046,1099,1129,1133,1211,1238,1251,1257,1334,1349,1355,1403,1410,1457,1464,1476,1483,1487,1491,1543,1555,1561,1566,1647,1668,1672,1676,1690 'docs/design':233 'docs/prd':225 'document':26,169,496,652,671,739,1102,1187,1313,1424,1492,1518,1550,1598,1784 'document-review':738,1549 'downstream':495 'duplic':422 'empti':434,444 'enabl':465,916,1764 'engin':4,22,537,547,1146,1270,1368 'ensur':672 'enter':146 'entri':571,576,587 'error':1803,1805 'evid':776 'exact':1106 'exclud':479 'execut':40,79,935 'exist':11,28,227,235,240,404,813,848,1140,1264,1362,1635,1688 'explicit':1835 'export':1069,1178,1203 'extern':548 'fail':1817 'failur':1819 'fals':658,1505 'featur':542,552,785,1137,1147,1261,1271,1359,1369,1797 'feedback':840,854,1680,1702 'fetch':1423 'file':565,570,590,1053,1067,1158,1176,1282,1380 'filter':130 'final':597,1776,1778 'find':752,1565,1808 'first':136 'flag':713,895,1743,1798,1829,1845 'flatten':409 'focus':375,378,455,769,1303,1414,1594 'follow':541,747,803,1136,1260,1358,1559 'forward':1050 'frontend':206,219,312,1038,1332,1340,1344,1347,1353,1660,1664,1666 'fulli':192 'fullstack':195,306,1010,1022,1024,1113,1218,1220,1452 'function':202,355,359 'gate':389,690 'generat':5,15,204,249,270,278,302,318,518,531,934,1047,1100,1127,1248,1346,1783,1816 'granular':978 'group':257,477,491 'guid':67 'handl':1804 'hexagon':184 'hierarchi':1417 'hint':402,1815 'human':187,261,461,897,912,1745,1760,1832,1836,1847 'id':416 'ident':35 'implement':648,977,1108,1486 'implementation-granular':976 'import':487 'includ':594,1782 'incomplet':446 'incorrect':490 'independ':665,673,1512 'indic':445 'initi':154 'input':96,721,1529 'interact':1421 'interfac':1056,1163,1166,1287,1290,1385,1388,1599 'intervent':1848 'inventori':1064,1171,1173,1195,1295,1297,1321,1393,1395,1432 'investig':582,604,708 'invoc':344,521,628,732,822,1117,1461,1540,1644 'invok':49,1225 'item':871,875,1719,1723,1793,1799 'json':726,1534 'layer':181,1308 'least':391,430 'leav':880,1728 'log':1818 'logic':1310 'loop':269,301,886,1734,1844 'major':870,1718 'manag':1419 'mandatori':1831 'map':286,472,953,969,1002,1039 'match':411 'maximum':888,1736 'minor':874,1722 'miss':419 'mode':544,760,846,1009,1023,1112,1143,1219,1267,1365,1447,1453,1580,1686 'modul':1161,1285,1383 'mvc':182 'n':114 'name':555,1149,1273,1371,1787 'need':243,807,1629 'next':99,110,920,923,1768,1771,1800 'non':433 'non-empti':432 'none':185 'note':209,659,1506 'noth':1809 'omit':660,1507 'one':74,105,392,431,510,800,1091 'oper':543,845,1142,1266,1364,1685 'orchestr':32,39,66,68,124 'otherwis':879,1727 'output':92,115,221,224,232,383,387,559,606,610,620,656,682,684,688,724,759,766,788,792,817,852,857,863,1042,1071,1075,1089,1153,1213,1217,1234,1277,1326,1330,1375,1406,1437,1441,1451,1496,1500,1503,1520,1524,1532,1572,1576,1579,1586,1611,1615,1639,1693,1697,1700,1705,1711,1780 'outsid':591 'overview':245 'parent':1182,1298,1396,1587,1602 'pass':53,112,256,914,1762 'path':55,165,366,369,612,622,653,662,1186,1302,1400,1493,1509,1591 'per':267,299,328,501,955,1080,1462,1541,1645 'per-unit':266,298,327,500,954,1079 'permit':61,70 'phase':135,143,246,274,315,928,930,1000 'point':463,489,572,577,584,588 'prd':6,172,175,223,228,248,259,317,324,339,517,528,532,538,553,561,566,573,599,611,621,639,645,651,743,748,755,772,829,833,836,849,994,997,1013,1027,1183,1185,1299,1301,1397,1399,1588,1590,1603 'prd-creator':527,828 'prdunit':408 'prerequisit':617 'present':466 'primari':1052,1157,1160,1281,1284,1379,1382 'proceed':395,696,921,1769 'process':73,128,330,503,957,1082 'produc':310 'profil':982 'project':1813 'prompt':357,533,641,744,834,1130,1252,1350,1478,1556,1673 'protocol':41 'provid':550,579 'public':1055,1068,1162,1165,1177,1286,1289,1384,1387 'publicexport':439 'qualiti':388,689 'ra':373 'rational':885,1733 're':293,449,705 're-discoveri':292 're-run':448,704 'recip':2 'recipe-reverse-engin':1 'recommend':190,778 'refer':179,370,1407 'regardless':899,1747 'regist':134,319,945 'registr':133 'reject':810,1632,1839 'relat':564,569 'relatedfil':460,1059 'report':59,1777,1779,1824 'request':280,941 'requir':95,210,720,872,973,1528,1720,1834 'resolut':777 'result':60,290,844,860,1684,1708 'reus':287 'reveal':589 'revers':3,21,536,546,1145,1269,1367 'reverse-engin':535,545,1144,1268,1366 'review':18,188,262,272,304,462,488,698,716,719,740,742,745,768,804,839,853,898,904,913,1527,1551,1553,1557,1593,1626,1679,1701,1746,1752,1761,1790,1833,1838 'revis':273,305,795,808,832,890,1618,1630,1670,1738,1842 'rout':436,1065,1174,1201 'run':450,706 'scope':157,252,285,325,340,351,356,360,452,482,549,580,593,668,679,952,968,1060,1101,1180,1515,1604 'scope-discover':350,451,678 'score':1789 'see':63 'separ':1458 'sequenti':81,509,1090,1229 'servic':1311 'set':413 'sever':867,1715 'shallow':710 'singl':255 'skill' 'skill-recipe-reverse-engineer' 'sourc':470 'source-shinpr' 'sourceunit':406,985 'specifi':381 'standard':1008,1111,1446 'start':108,583 'state':1418 'status':149,805,901,905,1627,1749,1753,1791 'step':75,80,90,100,103,113,140,152,250,263,281,288,295,322,331,337,385,497,515,557,608,613,618,624,654,686,717,722,728,757,764,790,793,815,850,855,861,918,943,948,958,964,1040,1073,1076,1087,1096,1151,1215,1275,1328,1373,1404,1439,1442,1449,1494,1498,1501,1522,1525,1530,1536,1570,1574,1577,1584,1613,1616,1624,1637,1691,1695,1698,1703,1709,1766,1801 'step_1_output.discoveredunits':975 'step_1_output.prdunits':403,467,508,984 'stop':1843 'store':382,605,683,787,1070,1212,1325,1436,1519,1610 'straightforward':878,1726 'structur':1814 'sub':51,121 'sub-ag':50,120 'subag':65,345,522,629,733,823,1118,1239,1335,1465,1544,1648 'subagents-orchestration-guid':64 'summari':1781,1826 'tabl':1785 'target':30,164,361,365,368,756,1048,1569 'task':132 'taskcreat':138,321,947 'taskupd':151 'technic':214,217,981,1020,1033,1036,1124,1245,1342,1595,1654,1662 'technical-design':213,1019,1032,1123,1244,1653 'technical-designer-frontend':216,1035,1341,1661 'technicalprofile.primarymodules':1051 'technicalprofile.publicinterfaces':1054 'test':1066,1175 'testfil':437 'thorough':602 'three':443 'time':78,514,1095 'tool':48,62,71,343,520,627,731,821,1116,1460,1539,1643 'topic-agent-skills' 'topic-agentic-ai' 'topic-ai-agents' 'topic-automation' 'topic-claude-code' 'topic-claude-code-plugin' 'topic-code-quality' 'topic-developer-tools' 'topic-development-workflow' 'topic-llm-orchestration' 'topic-productivity' 'topic-prompt-engineering' 'trace':586,987 'trigger':797,1620 'true':551 'type':346,523,630,650,734,754,824,1119,1240,1336,1466,1488,1545,1567,1649,1786 'ui':1420 'undocu':784,1796 'unifi':254 'unit':84,106,203,260,268,300,309,329,336,393,397,418,421,425,440,458,471,480,502,506,511,554,562,567,574,902,924,927,956,963,979,990,995,998,1006,1014,1028,1043,1063,1081,1085,1092,1148,1155,1159,1164,1168,1170,1172,1194,1224,1272,1279,1283,1288,1292,1294,1296,1320,1370,1377,1381,1386,1390,1392,1394,1431,1607,1750,1772,1775,1823 'unitinventori':427,1062 'updat':148,835,847,1674,1687 'use':137,150,159,320,578,946,974,983,1018,1193,1319,1430 'user':367,372,377,400,474,1811 'verbos':657,1504 'verif':16,271,303,616,674,725,751,763,775,843,859,1445,1533,1564,1583,1683,1707 'verifi':238,636,638,642,664,707,1448,1454,1472,1474,1479,1511 'verifiableclaimcount':694,702 'version':598 'within':82,141 'without':127 'work':45 'workflow':19,23,244,349,526,633,737,827,1122,1243,1339,1469,1548,1652,1659 'yes':189,197,199,307,1025,1221","prices":[{"id":"aafa7a89-feae-40d4-a91b-9cb59059b537","listingId":"59e1a3cf-b270-4cf3-b976-b7b322122e33","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"shinpr","category":"claude-code-workflows","install_from":"skills.sh"},"createdAt":"2026-04-18T22:03:15.118Z"}],"sources":[{"listingId":"59e1a3cf-b270-4cf3-b976-b7b322122e33","source":"github","sourceId":"shinpr/claude-code-workflows/recipe-reverse-engineer","sourceUrl":"https://github.com/shinpr/claude-code-workflows/tree/main/skills/recipe-reverse-engineer","isPrimary":false,"firstSeenAt":"2026-04-18T22:03:15.118Z","lastSeenAt":"2026-05-02T18:53:52.966Z"}],"details":{"listingId":"59e1a3cf-b270-4cf3-b976-b7b322122e33","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"shinpr","slug":"recipe-reverse-engineer","github":{"repo":"shinpr/claude-code-workflows","stars":327,"topics":["agent-skills","agentic-ai","ai-agents","automation","claude-code","claude-code-plugin","code-quality","developer-tools","development-workflow","llm-orchestration","productivity","prompt-engineering","skills"],"license":"mit","html_url":"https://github.com/shinpr/claude-code-workflows","pushed_at":"2026-05-02T15:39:17Z","description":"Production-ready development workflows for Claude Code, powered by specialized AI agents.","skill_md_sha":"5b9e9f33907e063081db48e10d5aa6d71829b6f1","skill_md_path":"skills/recipe-reverse-engineer/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/shinpr/claude-code-workflows/tree/main/skills/recipe-reverse-engineer"},"layout":"multi","source":"github","category":"claude-code-workflows","frontmatter":{"name":"recipe-reverse-engineer","description":"Generate PRD and Design Docs from existing codebase through discovery, generation, verification, and review workflow"},"skills_sh_url":"https://skills.sh/shinpr/claude-code-workflows/recipe-reverse-engineer"},"updatedAt":"2026-05-02T18:53:52.966Z"}}