{"id":"1cf834c8-72fc-4636-80cb-fbbeabb3941a","shortId":"Mzu2BB","kind":"skill","title":"calibrate","tagline":"Interactive calibration session to teach the clone how you think, decide, and prioritize. Run /calibrate to start a session, /calibrate status to see coverage, or /calibrate <domain> to focus on a specific area (e.g., /calibrate prioritization). Use when you want to improve clone","description":"# Calibrate -- Teach Your Clone How You Think\n\nInteractive scenario-based calibration sessions that extract your decision heuristics, values, and reasoning patterns. Each session takes 5-10 minutes and significantly improves how well the clone models your thinking.\n\n## How It Works\n\n1. **Gap Analysis** -- Check which domains of your thinking are under-modeled\n2. **Scenario Presentation** -- Present a realistic scenario that forces a judgment call\n3. **Adaptive Follow-ups** -- Probe deeper based on your initial response\n4. **Disagreement Probes** -- Challenge your answer to surface nuance and exceptions\n5. **Knowledge Storage** -- Extract and store decision patterns and values at high confidence\n\n## Commands\n\n- `/calibrate` -- Start a calibration session (auto-picks the least-covered domain)\n- `/calibrate status` -- Show calibration coverage and gaps\n- `/calibrate <domain>` -- Focus on a specific domain\n\n### Available Domains\n\ntech_decisions, communication, conflict, prioritization, leadership, quality, collaboration, risk, creativity, time_management\n\n## Session Protocol\n\nWhen the user invokes `/calibrate`, follow this exact protocol:\n\n### Phase 1: Status Check\n\n1. Call `user_model_recall` to load the current user model\n2. Analyze coverage across all calibration domains\n3. Show a brief status summary:\n\n   ```\n   Clone calibration: 42% [========------------]\n\n   Gaps: Tech Decisions (10%), Prioritization (0%), Risk (20%)\n   Best covered: Communication (80%), Quality (70%)\n\n   Starting session on: Prioritization (least covered)\n   ```\n\n### Phase 2: Scenario\n\n1. Present ONE scenario from the target domain\n2. Frame it conversationally: \"Here's a situation I'd like to understand how you'd handle...\"\n3. Wait for the user's response -- do NOT rush or provide options\n\n### Phase 3: Follow-up Probes\n\nAfter the user responds, ask 2-3 follow-up questions:\n\n1. **Clarification**: \"What's the main factor driving that choice?\"\n2. **Exception probe**: \"Would anything flip your decision? What would have to be different?\"\n3. **Disagreement probe**: \"I think based on what I know about you, you'd lean toward [opposite]. Am I wrong?\" (Only use this if you have a genuine hypothesis from the user model)\n\n### Phase 4: Extraction & Storage\n\nAfter the probing conversation, extract the knowledge:\n\n1. Identify the **decision pattern** (the heuristic or principle behind their choice)\n2. Identify any **values** revealed (what they prioritize and why)\n3. Identify any **exceptions** to the pattern\n\nStore these using `user_model_recall` confirmation -- tell the user what you learned:\n\n```\nHere's what I learned from this scenario:\n\nDecision pattern: \"Ship the working solution first, refactor later -- unless\ntech debt would block other teams\" (weight: 0.8)\n  Context: deadline pressure, stakeholder management\n  Exception: when the debt affects shared infrastructure\n\nValue: Pragmatism over perfectionism in time-constrained situations\n\nDoes this capture it accurately? Anything to adjust?\n```\n\n### Phase 5: User Correction\n\nIf the user corrects or refines your extraction:\n\n- Update the pattern/value immediately\n- Store with confidence 0.85 (explicit calibration = high confidence)\n- Thank them and note what you adjusted\n\n### Phase 6: Continue or Close\n\nAsk: \"Want to continue with another scenario, or is this a good stopping point?\"\n\nIf continuing, pick the next least-covered domain.\n\n## Important Rules\n\n- **One scenario at a time** -- never present multiple scenarios\n- **Wait for responses** -- don't anticipate or provide sample answers\n- **Be genuinely curious** -- these are real conversations, not quizzes\n- **Store at high confidence (0.85)** -- explicit calibration is the most reliable signal\n- **Show what you learned** -- always summarize extractions and ask for confirmation\n- **Track progress** -- update the coverage % as you go\n- **Respect time** -- if the user seems done, gracefully close even if there are more gaps\n- **No judgment** -- there are no right or wrong answers, only preferences to understand","tags":["calibrate","nomos","project-nomos","agent-memory","agent-skills","agentic-ai","ai-agents","ai-assistant","autonomous-agents","claude","claude-ai","claude-code"],"capabilities":["skill","source-project-nomos","skill-calibrate","topic-agent-memory","topic-agent-skills","topic-agentic-ai","topic-ai-agents","topic-ai-assistant","topic-autonomous-agents","topic-claude","topic-claude-ai","topic-claude-code","topic-claude-skills","topic-digital-clone","topic-llm"],"categories":["nomos"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/project-nomos/nomos/calibrate","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add project-nomos/nomos","source_repo":"https://github.com/project-nomos/nomos","install_from":"skills.sh"}},"qualityScore":"0.457","qualityRationale":"deterministic score 0.46 from registry signals: · indexed on github topic:agent-skills · 14 github stars · SKILL.md body (4,043 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":"2026-04-22T01:02:18.354Z","embedding":null,"createdAt":"2026-04-21T19:04:07.371Z","updatedAt":"2026-04-22T01:02:18.354Z","lastSeenAt":"2026-04-22T01:02:18.354Z","tsv":"'-10':70 '-3':302 '/calibrate':16,21,27,35,147,160,167,193 '0':234 '0.8':442 '0.85':491,565 '1':85,199,202,252,307,375 '10':232 '2':98,213,250,260,301,317,387 '20':236 '3':110,220,277,291,331,397 '4':122,365 '42':228 '5':69,133,473 '6':504 '70':242 '80':240 'accur':468 'across':216 'adapt':111 'adjust':471,502 'affect':452 'alway':577 'analysi':87 'analyz':214 'anoth':513 'answer':127,551,615 'anticip':547 'anyth':321,469 'area':33 'ask':300,508,581 'auto':153 'auto-pick':152 'avail':173 'base':54,117,336 'behind':384 'best':237 'block':438 'brief':223 'calibr':1,3,44,55,150,163,218,227,493,567 'call':109,203 'captur':466 'challeng':125 'check':88,201 'choic':316,386 'clarif':308 'clone':8,43,47,78,226 'close':507,600 'collabor':182 'command':146 'communic':177,239 'confid':145,490,495,564 'confirm':410,583 'conflict':178 'constrain':462 'context':443 'continu':505,511,523 'convers':263,371,558 'correct':475,479 'cover':158,238,248,529 'coverag':25,164,215,588 'creativ':184 'curious':554 'current':210 'd':269,275,344 'deadlin':444 'debt':436,451 'decid':12 'decis':60,139,176,231,324,378,425 'deeper':116 'differ':330 'disagr':123,332 'domain':90,159,172,174,219,259,530 'done':598 'drive':314 'e.g':34 'even':601 'exact':196 'except':132,318,400,448 'explicit':492,566 'extract':58,136,366,372,483,579 'factor':313 'first':431 'flip':322 'focus':29,168 'follow':113,194,293,304 'follow-up':112,292,303 'forc':106 'frame':261 'gap':86,166,229,606 'genuin':358,553 'go':591 'good':519 'grace':599 'handl':276 'heurist':61,381 'high':144,494,563 'hypothesi':359 'identifi':376,388,398 'immedi':487 'import':531 'improv':42,74 'infrastructur':454 'initi':120 'interact':2,51 'invok':192 'judgment':108,608 'know':340 'knowledg':134,374 'later':433 'leadership':180 'lean':345 'learn':416,421,576 'least':157,247,528 'least-cov':156,527 'like':270 'load':208 'main':312 'manag':186,447 'minut':71 'model':79,97,205,212,363,408 'multipl':540 'never':538 'next':526 'note':499 'nuanc':130 'one':254,533 'opposit':347 'option':289 'pattern':65,140,379,403,426 'pattern/value':486 'perfection':458 'phase':198,249,290,364,472,503 'pick':154,524 'point':521 'pragmat':456 'prefer':617 'present':100,101,253,539 'pressur':445 'principl':383 'priorit':14,36,179,233,246,394 'probe':115,124,295,319,333,370 'progress':585 'protocol':188,197 'provid':288,549 'qualiti':181,241 'question':306 'quizz':560 'real':557 'realist':103 'reason':64 'recal':206,409 'refactor':432 'refin':481 'reliabl':571 'respect':592 'respond':299 'respons':121,283,544 'reveal':391 'right':612 'risk':183,235 'rule':532 'run':15 'rush':286 'sampl':550 'scenario':53,99,104,251,255,424,514,534,541 'scenario-bas':52 'see':24 'seem':597 'session':4,20,56,67,151,187,244 'share':453 'ship':427 'show':162,221,573 'signal':572 'signific':73 'situat':267,463 'skill' 'skill-calibrate' 'solut':430 'source-project-nomos' 'specif':32,171 'stakehold':446 'start':18,148,243 'status':22,161,200,224 'stop':520 'storag':135,367 'store':138,404,488,561 'summar':578 'summari':225 'surfac':129 'take':68 'target':258 'teach':6,45 'team':440 'tech':175,230,435 'tell':411 'thank':496 'think':11,50,81,93,335 'time':185,461,537,593 'time-constrain':460 'topic-agent-memory' 'topic-agent-skills' 'topic-agentic-ai' 'topic-ai-agents' 'topic-ai-assistant' 'topic-autonomous-agents' 'topic-claude' 'topic-claude-ai' 'topic-claude-code' 'topic-claude-skills' 'topic-digital-clone' 'topic-llm' 'toward':346 'track':584 'under-model':95 'understand':272,619 'unless':434 'up':114 'updat':484,586 'use':37,352,406 'user':191,204,211,281,298,362,407,413,474,478,596 'valu':62,142,390,455 'wait':278,542 'want':40,509 'weight':441 'well':76 'work':84,429 'would':320,326,437 'wrong':350,614","prices":[{"id":"0d63ce40-5ef2-4195-8c73-0f891bf014a5","listingId":"1cf834c8-72fc-4636-80cb-fbbeabb3941a","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"project-nomos","category":"nomos","install_from":"skills.sh"},"createdAt":"2026-04-21T19:04:07.371Z"}],"sources":[{"listingId":"1cf834c8-72fc-4636-80cb-fbbeabb3941a","source":"github","sourceId":"project-nomos/nomos/calibrate","sourceUrl":"https://github.com/project-nomos/nomos/tree/main/skills/calibrate","isPrimary":false,"firstSeenAt":"2026-04-21T19:04:07.371Z","lastSeenAt":"2026-04-22T01:02:18.354Z"}],"details":{"listingId":"1cf834c8-72fc-4636-80cb-fbbeabb3941a","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"project-nomos","slug":"calibrate","github":{"repo":"project-nomos/nomos","stars":14,"topics":["agent-memory","agent-skills","agentic-ai","ai-agents","ai-assistant","autonomous-agents","claude","claude-ai","claude-code","claude-skills","digital-clone","llm","mcp","multi-agent","multi-agent-systems","ollama","self-hosted"],"license":"mit","html_url":"https://github.com/project-nomos/nomos","pushed_at":"2026-04-18T00:18:33Z","description":"Your AI digital clone — learns who you are, acts on your behalf, remembers everything. Persistent vector memory, multi-agent teams, 60+ skills, smart model routing. Self-hosted, encrypted, multi-provider (Claude/Ollama/OpenRouter). Deploy to Slack, Discord, Telegram, WhatsApp & more in minutes.","skill_md_sha":"addfc341d6bfd25bd1fcb7a7286c1cb2a841b29a","skill_md_path":"skills/calibrate/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/project-nomos/nomos/tree/main/skills/calibrate"},"layout":"multi","source":"github","category":"nomos","frontmatter":{"name":"calibrate","description":"Interactive calibration session to teach the clone how you think, decide, and prioritize. Run /calibrate to start a session, /calibrate status to see coverage, or /calibrate <domain> to focus on a specific area (e.g., /calibrate prioritization). Use when you want to improve clone fidelity, teach it your values, or help it understand your decision-making style."},"skills_sh_url":"https://skills.sh/project-nomos/nomos/calibrate"},"updatedAt":"2026-04-22T01:02:18.354Z"}}