{"id":"3d267504-2f6d-49c5-8dcf-d759918e921c","shortId":"nJv4c3","kind":"skill","title":"skill-design-guide","tagline":"Comprehensive guide for designing, building, testing, and distributing Claude skills. Use when user wants to \"create a skill\", \"build a skill\", \"write a SKILL.md\", \"design a workflow skill\", \"make an MCP skill\", or asks about skill structure, frontmatter, trigger phrases, or skil","description":"# Skill Design Guide\n\nInteractive guide for building Claude skills from scratch. Walks users through\nthe full lifecycle: use case definition, folder structure, frontmatter,\ninstructions, design patterns, testing, iteration, and distribution.\n\n## What Is a Skill\n\nA skill is a folder containing instructions that teach Claude how to handle\nspecific tasks or workflows. Instead of re-explaining preferences and processes\nevery conversation, skills teach Claude once.\n\n**Required file:** `SKILL.md` — main instructions in Markdown with YAML frontmatter.\n\n**Optional directories:**\n\n- `scripts/` — executable code (Python, Bash, etc.)\n- `references/` — documentation loaded as needed\n- `assets/` — templates, fonts, icons used in output\n\n**Progressive disclosure (three levels):**\n\n1. YAML frontmatter — always loaded in system prompt. Tells Claude when to use the skill.\n2. SKILL.md body — loaded when Claude thinks the skill is relevant. Full instructions.\n3. Linked files — scripts, references, assets. Loaded only as needed.\n\n**Three skill categories:**\n\n1. Document and Asset Creation — consistent output (docs, presentations, code, designs)\n2. Workflow Automation — multi-step processes with consistent methodology\n3. MCP Enhancement — workflow guidance on top of MCP tool access\n\n## Instructions\n\nGuide the user through each step interactively. Ask clarifying questions\nbefore moving to the next step. Do not rush — quality of each step matters\nmore than speed through the workflow.\n\n### Step 1: Define Use Cases\n\nAsk the user for 2-3 concrete use cases the skill should enable.\n\nFor each use case, capture:\n\n- **Trigger:** What the user would say (e.g., \"help me plan this sprint\")\n- **Steps:** What the workflow involves\n- **Tools:** Built-in Claude capabilities, MCP tools, or scripts needed\n- **Result:** What success looks like\n\nExample of a well-defined use case:\n\n```\nUse Case: Project Sprint Planning\nTrigger: User says \"help me plan this sprint\" or \"create sprint tasks\"\nSteps:\n1. Fetch current project status from Linear (via MCP)\n2. Analyze team capacity\n3. Suggest task prioritization\n4. Create tasks in Linear with proper labels and estimates\nResult: Fully planned sprint with tasks created\n```\n\nDetermine whether the skill is standalone or MCP-enhanced. Standalone skills\nuse Claude's built-in capabilities only. MCP-enhanced skills coordinate with\none or more MCP servers.\n\n### Step 2: Set Up Folder Structure\n\nCreate the skill folder following these rules:\n\n**Folder naming (kebab-case only):**\n\n- `notion-project-setup` — correct\n- `Notion Project Setup` — wrong (spaces, capitals)\n- `notion_project_setup` — wrong (underscores)\n- `NotionProjectSetup` — wrong (capitals)\n\n**Inside the folder:**\n\n- `SKILL.md` — required, exact casing (not SKILL.MD, skill.md, etc.)\n- `scripts/` — if executable code is needed\n- `references/` — if detailed documentation is needed\n- `assets/` — if templates or static files are needed\n- No `README.md` inside the skill folder (docs go in SKILL.md or references/)\n\n### Step 3: Write YAML Frontmatter\n\nThe frontmatter is how Claude decides whether to load the skill. This is the\nmost important part to get right.\n\n**Required format:**\n\n```yaml\n---\nname: your-skill-name\ndescription: What it does. Use when user asks to [specific phrases].\n---\n```\n\n**Required fields:**\n\n- `name` — kebab-case, no spaces or capitals, should match folder name\n- `description` — must include WHAT the skill does AND WHEN to use it, under 1024 characters, no XML angle brackets\n\n**Optional fields:**\n\n- `license` — e.g., MIT, Apache-2.0\n- `compatibility` — environment requirements (1-500 chars)\n- `metadata` — custom key-value pairs (author, version, mcp-server, etc.)\n\n**Security restrictions:**\n\n- No XML angle brackets in frontmatter\n- No \"claude\" or \"anthropic\" in the skill name (reserved)\n\nConsult `references/yaml-frontmatter-reference.md` for the complete field reference.\n\n### Step 4: Write the Description Field\n\nThe description determines whether the skill triggers correctly. Structure it as:\n\n```\n[What it does] + [When to use it] + [Key capabilities]\n```\n\n**Good descriptions:**\n\n```yaml\n# Specific and actionable\ndescription: Analyzes Figma design files and generates developer handoff\n  documentation. Use when user uploads .fig files, asks for \"design specs\",\n  \"component documentation\", or \"design-to-code handoff\".\n\n# Includes trigger phrases\ndescription: Manages Linear project workflows including sprint planning,\n  task creation, and status tracking. Use when user mentions \"sprint\",\n  \"Linear tasks\", \"project planning\", or asks to \"create tickets\".\n\n# Clear value proposition\ndescription: End-to-end customer onboarding workflow for PayFlow. Handles\n  account creation, payment setup, and subscription management. Use when\n  user says \"onboard new customer\", \"set up subscription\", or \"create\n  PayFlow account\".\n```\n\n**Bad descriptions:**\n\n```yaml\n# Too vague — won't trigger reliably\ndescription: Helps with projects.\n\n# Missing triggers — Claude won't know when to load it\ndescription: Creates sophisticated multi-page documentation systems.\n\n# Too technical, no user triggers\ndescription: Implements the Project entity model with hierarchical relationships.\n```\n\nIf there is risk of over-triggering, add negative triggers:\n\n```yaml\ndescription: Advanced data analysis for CSV files. Use for statistical\n  modeling, regression, clustering. Do NOT use for simple data exploration\n  (use data-viz skill instead).\n```\n\n### Step 5: Write Main Instructions\n\nAfter the frontmatter, write the instructions in Markdown. Follow this structure:\n\n```markdown\n# Your Skill Name\n\n## Instructions\n\n### Step 1: [First Major Step]\n\nClear explanation of what happens.\n\nExample:\npython scripts/fetch_data.py --project-id PROJECT_ID\nExpected output: [describe what success looks like]\n\n(More steps as needed)\n\n## Examples\n\n### Example 1: [Common scenario]\n\nUser says: \"Set up a new marketing campaign\"\nActions:\n\n1. Fetch existing campaigns via MCP\n2. Create new campaign with provided parameters\n   Result: Campaign created with confirmation link\n\n## Troubleshooting\n\n### Error: [Common error message]\n\nCause: [Why it happens]\nSolution: [How to fix]\n```\n\n**Best practices:**\n\n- Be specific and actionable — show exact commands, expected outputs\n- Include error handling for common failures\n- Reference bundled files clearly: \"Consult `references/api-patterns.md` for rate limiting guidance\"\n- Use numbered lists and bullet points over prose\n- Put critical instructions at the top with clear headers\n- Keep SKILL.md under 5,000 words — move detailed docs to `references/`\n\n**For critical validations**, consider bundling a script that performs checks\nprogrammatically rather than relying on language instructions. Code is\ndeterministic; language interpretation is not.\n\n### Step 6: Choose a Design Pattern\n\nHelp the user select the pattern that fits their use case:\n\n1. **Sequential Workflow Orchestration** — multi-step processes in a specific order with dependencies between steps\n2. **Multi-MCP Coordination** — workflows spanning multiple services (e.g., Figma to Drive to Linear to Slack)\n3. **Iterative Refinement** — output quality improves with validation and iteration loops\n4. **Context-Aware Tool Selection** — same outcome, different tools depending on context (file type, size, etc.)\n5. **Domain-Specific Intelligence** — specialized knowledge beyond tool access (compliance, regulations, best practices)\n\nConsult `references/design-patterns.md` for detailed examples and key techniques for each pattern.\n\n### Step 7: Test the Skill\n\nGuide the user through three testing areas:\n\n**Triggering tests:**\n\n- Does it load on obvious task requests?\n- Does it load on paraphrased requests?\n- Does it stay silent on unrelated topics?\n\nTest queries that should trigger:\n\n```\n\"Help me set up a new [Service] workspace\"\n\"I need to create a project in [Service]\"\n\"Initialize a [Service] project for Q4 planning\"\n```\n\nTest queries that should NOT trigger:\n\n```\n\"What's the weather?\"\n\"Help me write Python code\"\n\"Create a spreadsheet\" (unless the skill handles sheets)\n```\n\n**Functional tests:**\n\n- Valid outputs generated\n- API/MCP calls succeed\n- Error handling works\n- Edge cases covered\n\n**Performance comparison:**\nCompare with and without the skill on the same task. Track: number of messages,\nfailed API calls, tokens consumed, and whether user corrections were needed.\n\n**Debugging tip:** Ask Claude \"When would you use the [skill-name] skill?\"\nClaude will quote the description back — adjust based on what's missing.\n\n### Step 8: Iterate\n\nSkills are living documents. Watch for these signals:\n\n**Under-triggering** (skill doesn't load when it should):\n\n- Add more trigger phrases and keywords to the description\n- Include technical terms users might use\n- Check if the description is too narrow\n\n**Over-triggering** (skill loads for unrelated queries):\n\n- Add negative triggers (\"Do NOT use for...\")\n- Narrow the scope of the description\n- Clarify specific file types or services\n\n**Execution issues** (skill loads but results are inconsistent):\n\n- Improve instruction specificity\n- Add error handling\n- Use scripts for deterministic checks instead of language instructions\n- Put critical instructions at the top and repeat key points\n\n**Pro tip:** Iterate on a single challenging task until Claude succeeds, then\nextract the winning approach into the skill. This provides faster signal than\nbroad testing.\n\n### Step 9: Distribute\n\n**Claude.ai upload:**\n\n1. Zip the skill folder\n2. Settings then Capabilities then Skills then Upload\n3. Toggle the skill on\n\n**Claude Code:**\nPlace the skill folder in the Claude Code skills directory.\n\n**Organization-level deployment:**\nAdmins can deploy skills workspace-wide with automatic updates and centralized management.\n\n**GitHub (recommended for sharing):**\n\n- Public repo with clear README (repo-level, separate from skill folder)\n- Example usage and screenshots\n- Link from MCP documentation if applicable\n\n**API (programmatic use):**\n\n- `/v1/skills` endpoint for listing and managing skills\n- `container.skills` parameter in Messages API\n- Works with Claude Agent SDK for custom agents\n\nWhen writing about the skill, focus on outcomes not features:\n\n```\nGood: \"Set up complete project workspaces in seconds instead of 30 minutes\nof manual setup.\"\n\nBad: \"A folder containing YAML frontmatter and Markdown instructions that\ncalls our MCP server tools.\"\n```\n\n## References\n\nFor detailed information beyond these instructions:\n\n- `references/yaml-frontmatter-reference.md` — complete field reference with types, constraints, and annotated examples\n- `references/design-patterns.md` — all 5 design patterns with full example structures and key techniques\n- `references/troubleshooting.md` — common issues from upload errors to triggering problems to MCP failures\n- `references/quality-checklist.md` — pre-upload and post-upload validation checklist\n- `references/source/` — full original guide chapters (00 through 06) for authoritative deep reference","tags":["skill","design","guide","metaskills","agent-skills","ai-agents","anthropic","claude","claude-ai","claude-code","claude-code-skills","claude-desktop"],"capabilities":["skill","source-metaskills","skill-skill-design-guide","topic-agent-skills","topic-ai-agents","topic-anthropic","topic-claude","topic-claude-ai","topic-claude-code","topic-claude-code-skills","topic-claude-desktop","topic-claude-skills","topic-mcp","topic-skill-builder"],"categories":["skill-design-guide"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/metaskills/skill-design-guide","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add metaskills/skill-design-guide","source_repo":"https://github.com/metaskills/skill-design-guide","install_from":"skills.sh"}},"qualityScore":"0.454","qualityRationale":"deterministic score 0.45 from registry signals: · indexed on github topic:agent-skills · 9 github stars · SKILL.md body (11,037 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":"2026-05-18T19:08:47.238Z","embedding":null,"createdAt":"2026-05-18T13:14:30.574Z","updatedAt":"2026-05-18T19:08:47.238Z","lastSeenAt":"2026-05-18T19:08:47.238Z","tsv":"'-2.0':561 '-3':260 '-500':566 '/v1/skills':1451 '00':1567 '000':956 '06':1569 '1':146,187,251,332,565,834,864,876,1004,1375 '1024':549 '2':161,198,259,341,398,882,1020,1380 '3':174,208,345,479,1037,1388 '30':1491 '4':349,605,1048 '5':813,955,1065,1530 '6':988 '7':1091 '8':1242 '9':1371 'access':218,1074 'account':708,728 'action':635,875,913 'add':782,1262,1292,1322 'adjust':1235 'admin':1409 'advanc':787 'agent':1466,1470 'alway':149 'analysi':789 'analyz':342,637 'angl':553,584 'annot':1526 'anthrop':591 'apach':560 'api':1206,1448,1462 'api/mcp':1180 'applic':1447 'approach':1359 'area':1101 'ask':38,227,255,518,652,690,1218 'asset':135,179,190,458 'author':574 'authorit':1571 'autom':200 'automat':1417 'awar':1051 'back':1234 'bad':729,1496 'base':1236 'bash':128 'best':908,1077 'beyond':1072,1515 'bodi':163 'bracket':554,585 'broad':1368 'build':9,23,53 'built':292,382 'built-in':291,381 'bullet':939 'bundl':926,967 'call':1181,1207,1506 'campaign':874,879,885,890 'capabl':295,384,629,1383 'capac':344 'capit':426,434,531 'captur':272 'case':65,254,263,271,313,315,414,441,527,1003,1187 'categori':186 'caus':900 'central':1420 'challeng':1350 'chapter':1566 'char':567 'charact':550 'check':972,1277,1329 'checklist':1561 'choos':989 'clarifi':228,1305 'claud':13,54,90,110,155,166,294,379,487,589,744,1219,1229,1353,1393,1401,1465 'claude.ai':1373 'clear':694,838,928,950,1429 'cluster':798 'code':126,196,449,662,980,1166,1394,1402 'command':916 'common':865,897,923,1541 'compar':1191 'comparison':1190 'compat':562 'complet':601,1484,1519 'complianc':1075 'compon':656 'comprehens':5 'concret':261 'confirm':893 'consid':966 'consist':192,206 'constraint':1524 'consult':597,929,1079 'consum':1209 'contain':86,1499 'container.skills':1458 'context':1050,1060 'context-awar':1049 'convers':107 'coordin':390,1024 'correct':420,617,1213 'cover':1188 'creat':20,328,350,365,403,692,726,753,883,891,1140,1167 'creation':191,676,709 'critic':944,964,1335 'csv':791 'current':334 'custom':569,702,721,1469 'data':788,804,808 'data-viz':807 'debug':1216 'decid':488 'deep':1572 'defin':252,311 'definit':66 'depend':1017,1058 'deploy':1408,1411 'describ':853 'descript':511,536,608,611,631,636,667,697,730,738,752,765,786,1233,1270,1280,1304 'design':3,8,29,48,71,197,639,654,660,991,1531 'design-to-cod':659 'detail':454,959,1082,1513 'determin':366,612 'determinist':982,1328 'develop':643 'differ':1056 'directori':123,1404 'disclosur':143 'distribut':12,76,1372 'doc':194,472,960 'document':131,188,455,645,657,758,1247,1445 'doesn':1256 'domain':1067 'domain-specif':1066 'drive':1032 'e.g':279,558,1029 'edg':1186 'enabl':267 'end':699,701 'end-to-end':698 'endpoint':1452 'enhanc':210,375,388 'entiti':769 'environ':563 'error':896,898,920,1183,1323,1545 'estim':358 'etc':129,445,579,1064 'everi':106 'exact':440,915 'exampl':306,843,862,863,1083,1438,1527,1535 'execut':125,448,1311 'exist':878 'expect':851,917 'explain':102 'explan':839 'explor':805 'extract':1356 'fail':1205 'failur':924,1551 'faster':1365 'featur':1480 'fetch':333,877 'field':523,556,602,609,1520 'fig':650 'figma':638,1030 'file':113,176,463,640,651,792,927,1061,1307 'first':835 'fit':1000 'fix':907 'focus':1476 'folder':67,85,401,406,410,437,471,534,1379,1398,1437,1498 'follow':407,825 'font':137 'format':504 'frontmatt':42,69,121,148,482,484,587,819,1501 'full':62,172,1534,1563 'fulli':360 'function':1175 'generat':642,1179 'get':501 'github':1422 'go':473 'good':630,1481 'guid':4,6,49,51,220,1095,1565 'guidanc':212,934 'handl':93,707,921,1173,1184,1324 'handoff':644,663 'happen':842,903 'header':951 'help':280,322,739,993,1129,1162 'hierarch':772 'icon':138 'id':848,850 'implement':766 'import':498 'improv':1042,1319 'includ':538,664,672,919,1271 'inconsist':1318 'inform':1514 'initi':1145 'insid':435,468 'instead':98,811,1330,1489 'instruct':70,87,116,173,219,816,822,832,945,979,1320,1333,1336,1504,1517 'intellig':1069 'interact':50,226 'interpret':984 'involv':289 'issu':1312,1542 'iter':74,1038,1046,1243,1346 'kebab':413,526 'kebab-cas':412,525 'keep':952 'key':571,628,1085,1342,1538 'key-valu':570 'keyword':1267 'know':747 'knowledg':1071 'label':356 'languag':978,983,1332 'level':145,1407,1433 'licens':557 'lifecycl':63 'like':305,857 'limit':933 'linear':338,353,669,685,1034 'link':175,894,1442 'list':937,1454 'live':1246 'load':132,150,164,180,491,750,1106,1113,1258,1288,1314 'look':304,856 'loop':1047 'main':115,815 'major':836 'make':33 'manag':668,714,1421,1456 'manual':1494 'markdown':118,824,828,1503 'market':873 'match':533 'matter':243 'mcp':35,209,216,296,340,374,387,395,577,881,1023,1444,1508,1550 'mcp-enhanc':373,386 'mcp-server':576 'mention':683 'messag':899,1204,1461 'metadata':568 'methodolog':207 'might':1275 'minut':1492 'miss':742,1240 'mit':559 'model':770,796 'move':231,958 'multi':202,756,1009,1022 'multi-mcp':1021 'multi-pag':755 'multi-step':201,1008 'multipl':1027 'must':537 'name':411,506,510,524,535,595,831,1227 'narrow':1283,1299 'need':134,183,300,451,457,465,861,1138,1215 'negat':783,1293 'new':720,872,884,1134 'next':234 'notion':417,421,427 'notion-project-setup':416 'notionprojectsetup':432 'number':936,1202 'obvious':1108 'onboard':703,719 'one':392 'option':122,555 'orchestr':1007 'order':1015 'organ':1406 'organization-level':1405 'origin':1564 'outcom':1055,1478 'output':141,193,852,918,1040,1178 'over-trigg':779,1284 'page':757 'pair':573 'paramet':888,1459 'paraphras':1115 'part':499 'pattern':72,992,998,1089,1532 'payflow':706,727 'payment':710 'perform':971,1189 'phrase':44,521,666,1265 'place':1395 'plan':282,318,324,361,674,688,1151 'point':940,1343 'post':1558 'post-upload':1557 'practic':909,1078 'pre':1554 'pre-upload':1553 'prefer':103 'present':195 'priorit':348 'pro':1344 'problem':1548 'process':105,204,1011 'programmat':973,1449 'progress':142 'project':316,335,418,422,428,670,687,741,768,847,849,1142,1148,1485 'project-id':846 'prompt':153 'proper':355 'proposit':696 'prose':942 'provid':887,1364 'public':1426 'put':943,1334 'python':127,844,1165 'q4':1150 'qualiti':239,1041 'queri':1125,1153,1291 'question':229 'quot':1231 'rate':932 'rather':974 're':101 're-explain':100 'readm':1430 'readme.md':467 'recommend':1423 'refer':130,178,452,477,603,925,962,1511,1521,1573 'references/api-patterns.md':930 'references/design-patterns.md':1080,1528 'references/quality-checklist.md':1552 'references/source':1562 'references/troubleshooting.md':1540 'references/yaml-frontmatter-reference.md':598,1518 'refin':1039 'regress':797 'regul':1076 'relationship':773 'relev':171 'reli':976 'reliabl':737 'repeat':1341 'repo':1427,1432 'repo-level':1431 'request':1110,1116 'requir':112,439,503,522,564 'reserv':596 'restrict':581 'result':301,359,889,1316 'right':502 'risk':777 'rule':409 'rush':238 'say':278,321,718,868 'scenario':866 'scope':1301 'scratch':57 'screenshot':1441 'script':124,177,299,446,969,1326 'scripts/fetch_data.py':845 'sdk':1467 'second':1488 'secur':580 'select':996,1053 'separ':1434 'sequenti':1005 'server':396,578,1509 'servic':1028,1135,1144,1147,1310 'set':399,722,869,1131,1381,1482 'setup':419,423,429,711,1495 'share':1425 'sheet':1174 'show':914 'signal':1251,1366 'silent':1120 'simpl':803 'singl':1349 'size':1063 'skil':46 'skill':2,14,22,25,32,36,40,47,55,80,82,108,160,169,185,265,369,377,389,405,470,493,509,541,594,615,810,830,1094,1172,1196,1226,1228,1244,1255,1287,1313,1362,1378,1385,1391,1397,1403,1412,1436,1457,1475 'skill-design-guid':1 'skill-nam':1225 'skill-skill-design-guide' 'skill.md':28,114,162,438,443,444,475,953 'slack':1036 'solut':904 'sophist':754 'source-metaskills' 'space':425,529 'span':1026 'spec':655 'special':1070 'specif':94,520,633,911,1014,1068,1306,1321 'speed':246 'spreadsheet':1169 'sprint':284,317,326,329,362,673,684 'standalon':371,376 'static':462 'statist':795 'status':336,678 'stay':1119 'step':203,225,235,242,250,285,331,397,478,604,812,833,837,859,987,1010,1019,1090,1241,1370 'structur':41,68,402,618,827,1536 'subscript':713,724 'succeed':1182,1354 'success':303,855 'suggest':346 'system':152,759 'task':95,330,347,351,364,675,686,1109,1200,1351 'teach':89,109 'team':343 'technic':761,1272 'techniqu':1086,1539 'tell':154 'templat':136,460 'term':1273 'test':10,73,1092,1100,1103,1124,1152,1176,1369 'think':167 'three':144,184,1099 'ticket':693 'tip':1217,1345 'toggl':1389 'token':1208 'tool':217,290,297,1052,1057,1073,1510 'top':214,948,1339 'topic':1123 'topic-agent-skills' 'topic-ai-agents' 'topic-anthropic' 'topic-claude' 'topic-claude-ai' 'topic-claude-code' 'topic-claude-code-skills' 'topic-claude-desktop' 'topic-claude-skills' 'topic-mcp' 'topic-skill-builder' 'track':679,1201 'trigger':43,273,319,616,665,736,743,764,781,784,1102,1128,1157,1254,1264,1286,1294,1547 'troubleshoot':895 'type':1062,1308,1523 'under-trigg':1252 'underscor':431 'unless':1170 'unrel':1122,1290 'updat':1418 'upload':649,1374,1387,1544,1555,1559 'usag':1439 'use':15,64,139,158,253,262,270,312,314,378,515,546,626,646,680,715,793,801,806,935,1002,1223,1276,1297,1325,1450 'user':17,59,222,257,276,320,517,648,682,717,763,867,995,1097,1212,1274 'vagu':733 'valid':965,1044,1177,1560 'valu':572,695 'version':575 'via':339,880 'viz':809 'walk':58 'want':18 'watch':1248 'weather':1161 'well':310 'well-defin':309 'whether':367,489,613,1211 'wide':1415 'win':1358 'without':1194 'won':734,745 'word':957 'work':1185,1463 'workflow':31,97,199,211,249,288,671,704,1006,1025 'workspac':1136,1414,1486 'workspace-wid':1413 'would':277,1221 'write':26,480,606,814,820,1164,1472 'wrong':424,430,433 'xml':552,583 'yaml':120,147,481,505,632,731,785,1500 'your-skill-nam':507 'zip':1376","prices":[{"id":"2321b036-ab46-4dad-ad58-41952bea6f49","listingId":"3d267504-2f6d-49c5-8dcf-d759918e921c","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"metaskills","category":"skill-design-guide","install_from":"skills.sh"},"createdAt":"2026-05-18T13:14:30.574Z"}],"sources":[{"listingId":"3d267504-2f6d-49c5-8dcf-d759918e921c","source":"github","sourceId":"metaskills/skill-design-guide","sourceUrl":"https://github.com/metaskills/skill-design-guide","isPrimary":false,"firstSeenAt":"2026-05-18T13:14:30.574Z","lastSeenAt":"2026-05-18T19:08:47.238Z"}],"details":{"listingId":"3d267504-2f6d-49c5-8dcf-d759918e921c","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"metaskills","slug":"skill-design-guide","github":{"repo":"metaskills/skill-design-guide","stars":9,"topics":["agent-skills","ai-agents","anthropic","claude","claude-ai","claude-code","claude-code-skills","claude-desktop","claude-skills","mcp","skill-builder"],"license":null,"html_url":"https://github.com/metaskills/skill-design-guide","pushed_at":"2026-02-14T00:28:26Z","description":"Claude skill for building Claude skills. An interactive 9-step workflow distilled from Anthropic's complete guide — covering use cases, frontmatter, instructions, design patterns, testing, and distribution.","skill_md_sha":"f05b35061d71a033078aa33937b70745158251b2","skill_md_path":"SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/metaskills/skill-design-guide"},"layout":"root","source":"github","category":"skill-design-guide","frontmatter":{"name":"skill-design-guide","description":"Comprehensive guide for designing, building, testing, and distributing Claude skills. Use when user wants to \"create a skill\", \"build a skill\", \"write a SKILL.md\", \"design a workflow skill\", \"make an MCP skill\", or asks about skill structure, frontmatter, trigger phrases, or skill best practices. Walks through the full lifecycle from use case definition to distribution."},"skills_sh_url":"https://skills.sh/metaskills/skill-design-guide"},"updatedAt":"2026-05-18T19:08:47.238Z"}}