{"id":"3b4610a7-ca80-4b88-b04c-61d82957a02f","shortId":"8JAkSu","kind":"skill","title":"perplexity","tagline":"Web search and research using Perplexity AI. Use when user says \"search\", \"find\", \"look up\", \"ask\", \"research\", or \"what's the latest\" for generic queries. NOT for library/framework docs (use Context7) or workspace questions.","description":"# Perplexity Tools\n\nUse ONLY when user says \"search\", \"find\", \"look up\", \"ask\", \"research\", or \"what's the latest\" for generic queries. NOT for library/framework docs (use Context7), gt CLI (use Graphite MCP), or workspace questions (use Nx MCP).\n\n## Quick Reference\n\n**Which Perplexity tool?**\n- Need search results/URLs? → **Perplexity Search**\n- Need conversational answer? → **Perplexity Ask**\n- Need deep research? → **Researcher agent** (`/research <topic>`)\n\n**NOT Perplexity - use these instead:**\n- Library/framework docs → **Context7 MCP**\n- Graphite `gt` CLI → **Graphite MCP**\n- THIS workspace → **Nx MCP**\n- Specific URL → **URL Crawler**\n\n## Perplexity Search\n\n**When to use:**\n- Generic searches, finding resources\n- Current best practices, recent information\n- Tutorial/blog post discovery\n- User says \"search for...\", \"find...\", \"look up...\"\n\n**Default parameters (ALWAYS USE):**\n```typescript\nmcp__perplexity__perplexity_search({\n  query: \"your search query\",\n  max_results: 3,           // Default is 10 - too many!\n  max_tokens_per_page: 512  // Reduce per-result content\n})\n```\n\n**When to increase limits:**\nOnly if:\n- User explicitly needs comprehensive results\n- Initial search found nothing useful\n- Complex topic needs multiple sources\n\n```typescript\n// Increased limits (use sparingly)\nmcp__perplexity__perplexity_search({\n  query: \"complex topic\",\n  max_results: 5,\n  max_tokens_per_page: 1024\n})\n```\n\n## Perplexity Ask\n\n**When to use:**\n- Need conversational explanation, not search results\n- Synthesize information from web\n- Explain concepts with current context\n\n**Usage:**\n```typescript\nmcp__perplexity__perplexity_ask({\n  messages: [\n    {\n      role: \"user\",\n      content: \"Explain how postgres advisory locks work\"\n    }\n  ]\n})\n```\n\n**NOT for:**\n- Library documentation (use Context7)\n- Deep multi-source research (use researcher agent)\n\n## Prohibited Tool\n\n**NEVER use:** `mcp__perplexity__perplexity_research`\n\n**Use instead:** Researcher agent (`/research <topic>`)\n- Token cost: 30-50k tokens\n- Provides multi-source synthesis with citations\n- Use sparingly for complex questions only\n\n## Tool Selection Chain\n\n**Priority order:**\n1. **Context7 MCP** - Library/framework docs\n2. **Graphite MCP** - Any `gt` CLI mention\n3. **Nx MCP** - THIS workspace questions\n4. **Perplexity Search** - Generic searches\n5. **Perplexity Ask** - Conversational answers\n6. **Researcher agent** - Deep multi-source research\n7. **WebSearch** - Last resort (after Perplexity exhausted)\n\n## Examples\n\n**✅ CORRECT - Use Perplexity Search:**\n- \"Find postgres migration best practices\"\n- \"Search for React testing tutorials\"\n- \"Look up latest trends in microservices\"\n\n**✅ CORRECT - Use Perplexity Ask:**\n- \"Explain how postgres advisory locks work\"\n- \"What are the trade-offs of microservices?\"\n\n**❌ WRONG - Use Context7 instead:**\n- \"Search for React hooks documentation\" → Context7 MCP\n- \"Find Next.js routing docs\" → Context7 MCP\n- \"Look up Temporal workflow API\" → Context7 MCP\n\n**❌ WRONG - Use Graphite MCP instead:**\n- \"Search for gt stack commands\" → Graphite MCP\n- \"Find gt branch workflow\" → Graphite MCP\n\n**❌ WRONG - Use Nx MCP instead:**\n- \"Search for build config\" (in THIS workspace) → Nx MCP\n- \"Find project dependencies\" (in THIS workspace) → Nx MCP\n\n## Key Points\n\n- **Default to limited results** - avoid context bloat\n- **Library docs = Context7** - ALWAYS try Context7 first\n- **\"gt\" = Graphite MCP** - ANY \"gt\" mention uses Graphite\n- **Deep research = /research** - NOT perplexity_research tool\n- **Fallback chain** - Search → Ask → WebSearch (last resort)","tags":["perplexity","agent","toolkit","softaworks","agent-skills","automation","claude","claude-code","coding-agent","development"],"capabilities":["skill","source-softaworks","skill-perplexity","topic-agent-skills","topic-automation","topic-claude","topic-claude-code","topic-coding-agent","topic-development"],"categories":["agent-toolkit"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/softaworks/agent-toolkit/perplexity","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add softaworks/agent-toolkit","source_repo":"https://github.com/softaworks/agent-toolkit","install_from":"skills.sh"}},"qualityScore":"0.700","qualityRationale":"deterministic score 0.70 from registry signals: · indexed on github topic:agent-skills · 1689 github stars · SKILL.md body (3,592 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":"2026-05-03T00:52:50.346Z","embedding":null,"createdAt":"2026-04-18T21:54:40.926Z","updatedAt":"2026-05-03T00:52:50.346Z","lastSeenAt":"2026-05-03T00:52:50.346Z","tsv":"'-50':279 '/research':94,275,472 '1':300 '10':159 '1024':212 '2':305 '3':156,312 '30':278 '4':318 '5':207,323 '512':166 '6':328 '7':336 'advisori':246,371 'agent':93,262,274,330 'ai':8 'alway':143,458 'answer':86,327 'api':403 'ask':17,47,88,214,238,325,367,480 'avoid':452 'best':127,351 'bloat':454 'branch':420 'build':431 'chain':297,478 'citat':288 'cli':64,106,310 'command':415 'complex':188,203,292 'comprehens':181 'concept':229 'config':432 'content':171,242 'context':232,453 'context7':32,62,102,254,301,384,391,397,404,457,460 'convers':85,219,326 'correct':344,364 'cost':277 'crawler':116 'current':126,231 'deep':90,255,331,470 'default':141,157,448 'depend':440 'discoveri':133 'doc':30,60,101,304,396,456 'document':252,390 'exampl':343 'exhaust':342 'explain':228,243,368 'explan':220 'explicit':179 'fallback':477 'find':14,44,124,138,348,393,418,438 'first':461 'found':185 'generic':25,55,122,321 'graphit':66,104,107,306,408,416,422,463,469 'gt':63,105,309,413,419,462,466 'hook':389 'increas':174,194 'inform':130,225 'initi':183 'instead':99,272,385,410,428 'k':280 'key':446 'last':338,482 'latest':23,53,360 'librari':251,455 'library/framework':29,59,100,303 'limit':175,195,450 'lock':247,372 'look':15,45,139,358,399 'mani':161 'max':154,162,205,208 'mcp':67,73,103,108,112,146,198,235,267,302,307,314,392,398,405,409,417,423,427,437,445,464 'mention':311,467 'messag':239 'microservic':363,381 'migrat':350 'multi':257,284,333 'multi-sourc':256,283,332 'multipl':191 'need':79,84,89,180,190,218 'never':265 'next.js':394 'noth':186 'nx':72,111,313,426,436,444 'off':379 'order':299 'page':165,211 'paramet':142 'per':164,169,210 'per-result':168 'perplex':1,7,36,77,82,87,96,117,147,148,199,200,213,236,237,268,269,319,324,341,346,366,474 'point':447 'post':132 'postgr':245,349,370 'practic':128,352 'prioriti':298 'prohibit':263 'project':439 'provid':282 'queri':26,56,150,153,202 'question':35,70,293,317 'quick':74 'react':355,388 'recent':129 'reduc':167 'refer':75 'research':5,18,48,91,92,259,261,270,273,329,335,471,475 'resort':339,483 'resourc':125 'result':155,170,182,206,223,451 'results/urls':81 'role':240 'rout':395 'say':12,42,135 'search':3,13,43,80,83,118,123,136,149,152,184,201,222,320,322,347,353,386,411,429,479 'select':296 'skill' 'skill-perplexity' 'sourc':192,258,285,334 'source-softaworks' 'spare':197,290 'specif':113 'stack':414 'synthes':224 'synthesi':286 'tempor':401 'test':356 'token':163,209,276,281 'tool':37,78,264,295,476 'topic':189,204 'topic-agent-skills' 'topic-automation' 'topic-claude' 'topic-claude-code' 'topic-coding-agent' 'topic-development' 'trade':378 'trade-off':377 'trend':361 'tri':459 'tutori':357 'tutorial/blog':131 'typescript':145,193,234 'url':114,115 'usag':233 'use':6,9,31,38,61,65,71,97,121,144,187,196,217,253,260,266,271,289,345,365,383,407,425,468 'user':11,41,134,178,241 'web':2,227 'websearch':337,481 'work':248,373 'workflow':402,421 'workspac':34,69,110,316,435,443 'wrong':382,406,424","prices":[{"id":"67e700a2-5e6b-4b42-b45c-6a4b1f2be3ef","listingId":"3b4610a7-ca80-4b88-b04c-61d82957a02f","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"softaworks","category":"agent-toolkit","install_from":"skills.sh"},"createdAt":"2026-04-18T21:54:40.926Z"}],"sources":[{"listingId":"3b4610a7-ca80-4b88-b04c-61d82957a02f","source":"github","sourceId":"softaworks/agent-toolkit/perplexity","sourceUrl":"https://github.com/softaworks/agent-toolkit/tree/main/skills/perplexity","isPrimary":false,"firstSeenAt":"2026-04-18T21:54:40.926Z","lastSeenAt":"2026-05-03T00:52:50.346Z"}],"details":{"listingId":"3b4610a7-ca80-4b88-b04c-61d82957a02f","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"softaworks","slug":"perplexity","github":{"repo":"softaworks/agent-toolkit","stars":1689,"topics":["agent-skills","ai","automation","claude","claude-code","coding-agent","development"],"license":"mit","html_url":"https://github.com/softaworks/agent-toolkit","pushed_at":"2026-03-05T16:46:24Z","description":"A curated collection of skills for AI coding agents. Skills are packaged instructions and scripts that extend agent capabilities across development, documentation, planning, and professional workflows.","skill_md_sha":"dbf0d69fc080077152107af7d0dee139f728d1dd","skill_md_path":"skills/perplexity/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/softaworks/agent-toolkit/tree/main/skills/perplexity"},"layout":"multi","source":"github","category":"agent-toolkit","frontmatter":{"name":"perplexity","description":"Web search and research using Perplexity AI. Use when user says \"search\", \"find\", \"look up\", \"ask\", \"research\", or \"what's the latest\" for generic queries. NOT for library/framework docs (use Context7) or workspace questions."},"skills_sh_url":"https://skills.sh/softaworks/agent-toolkit/perplexity"},"updatedAt":"2026-05-03T00:52:50.346Z"}}