[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6P38KG60SBQMP44FDDAHERH":3,"skill-same:01M6P38KG60SBQMP44FDDAHERH":45},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":14,"categories":15,"keywords":16,"publishedAt":20,"linkOnly":21,"markdownContent":22,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":12,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":29,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M6P38KG60SBQMP44FDDAHERH","karpathy-guidelines",null,"dexter-cnx\u002Fagent-engineering-toolkit","dexter-cnx","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F73522389?v=4","01M6K09WYAAXJXMR26H00WJ2AT","01M6P370Z0XCZ7NC67MJ2MRW94",0,"MIT","https:\u002F\u002Fgithub.com\u002Fdexter-cnx\u002Fagent-engineering-toolkit\u002Fblob\u002Fmain\u002Foverlays\u002Fagent-karpathy\u002Fskills\u002Fkarpathy-guidelines\u002FSKILL.md",[],[17,18,19],"guidelines","karpathy guidelines","karpathy","2026-11-10 10:12:30",false,"# karpathy-guidelines\n\n## Purpose\nApply the Karpathy Layer V2 evaluation criteria to a skill or prompt to assess its quality across seven weighted dimensions before submitting it for automated optimization.\n\n## Use when\n- You are writing or reviewing a new skill and want a pre-flight quality check before running the automated eval pipeline\n- You need to explain why a skill scored low on a specific dimension and propose targeted improvements\n- You are onboarding a contributor and need to teach the evaluation model by example\n- You want to triage a low-scoring skill (below 0.60) before investing in a full mutation cycle\n\n## Do NOT use when\n- You only need a numeric score — use `runners\u002Feval_runner.py` directly — use a more specific skill instead.\n- You want to generate mutation candidates — use the `karpathy-mutation` skill instead — use a more specific skill instead.\n- You are checking for regression safety — use the `karpathy-regression` skill instead — use a more specific skill instead.\n- The skill is already above 0.85 and no specific dimension concern has been raised\n\n## Inputs required\n- The full SKILL.md content (all 13 sections)\n- The dimension weights from `evals\u002Frubrics\u002Fcoding_task_rubric.v1.json`\n- The specific concern or dimension the reviewer wants to focus on (optional)\n\n## Constraints\n- Apply the weights exactly: correctness 0.30, scope_discipline 0.15, simplicity 0.15, verifiability 0.15, architecture_alignment 0.10, token_efficiency 0.10, docs_hygiene 0.05\n- Do not substitute subjective opinion for the rubric criteria\n- Do not score a dimension higher than 0.8 if the Validation checklist has no executable items\n- Do not score docs_hygiene above 0.6 if any placeholder text (TODO, TBD, FIXME, {{}}) exists\n\n## Step-by-step workflow\n1. Read all 13 sections of the skill and confirm none are empty or contain placeholder text.\n2. Score `correctness` (0.30): check that workflow steps are ordered correctly, technically accurate, and the output contract matches the stated purpose.\n3. Score `scope_discipline` (0.15): count bullets in \"Use when\" and \"Do NOT use when\" — each section must have at least two items that cover distinct cases.\n4. Score `simplicity` (0.15): count numbered steps and total word count — flag any step containing \" and \" as a compound step, and any filler phrases.\n5. Score `verifiability` (0.15): read the Validation checklist — each bullet must contain an action verb (run, check, grep, confirm, assert, verify).\n6. Score `architecture_alignment` (0.10): check whether any steps recommend placing logic in the wrong layer (e.g., SDK calls in presentation widgets).\n7. Score `token_efficiency` (0.10): flag repeated content, filler sentences (\"It is important to note that…\"), and unnecessary preamble.\n8. Score `docs_hygiene` (0.05): confirm all 13 sections are present, no forbidden strings, and References section links to real files.\n9. Compute weighted final_score = sum(weight_i × score_i).\n10. List the two lowest-scoring dimensions and propose one concrete improvement for each.\n\n## Output contract\n- A score table with one row per dimension (name, weight, score, rationale in one sentence)\n- A computed final_score (rounded to 4 decimal places)\n- A binary checks summary (pass\u002Ffail for each of the 6 checks)\n- A prioritised improvement list (at most 3 items, highest ROI first)\n\n## Validation checklist\n- Confirm all 7 dimension scores are in [0.0, 1.0]\n- Confirm final_score = sum(weight × score) for all 7 dimensions\n- Confirm at least one improvement recommendation is actionable (not vague)\n- Confirm binary checks match the actual skill content (not assumed)\n\n## Related skills\n- `karpathy-mutation` — generate candidates based on low-scoring dimensions\n- `karpathy-regression` — check candidates for structural integrity\n- `karpathy-optimizer` — apply token policy and select the best candidate\n\n## References\n- [`..\u002F..\u002Fevals\u002Frubrics\u002Fcoding_task_rubric.v1.json`](..\u002F..\u002Fevals\u002Frubrics\u002Fcoding_task_rubric.v1.json)\n- [`..\u002F..\u002Fdocs\u002Feval-system.md`](..\u002F..\u002Fdocs\u002Feval-system.md)\n- [`..\u002F..\u002Fexamples\u002Fflutter_deeplink_full_cycle.md`](..\u002F..\u002Fexamples\u002Fflutter_deeplink_full_cycle.md)\n\n## Real example\nHistorical static snapshot: a reviewer applies this skill to `flutter-go-router-deeplink-wireup\u002FSKILL.md`. The Validation checklist has three bullets, two of which say \"ensure deep links work\" and \"make sure tests pass\" — both are vague. `verifiability` scores 0.40. The improvement recommendation is: \"Replace 'ensure deep links work' with 'Run `flutter test test\u002Frouting\u002Fdeep_link_test.dart` and confirm exit code 0'.\"\n\n## Real file output sample\n```text\ndocs\u002Feval-sessions\u002Fflutter-deeplink-manual-eval.md\nreports\u002Flatest_report.md\n```\n",{},"main","3035171e2986d5fdb445e694b9013fbdf58b3ff6","overlays\u002Fagent-karpathy\u002Fskills\u002Fkarpathy-guidelines","https:\u002F\u002Fgithub.com\u002Fdexter-cnx\u002Fagent-engineering-toolkit","https:\u002F\u002Fgithub.com\u002Fdexter-cnx","https:\u002F\u002Fdexter-cnx.github.io\u002Fportfolio\u002F","2026-05-26 14:24:27",{"source":7,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add dexter-cnx\u002Fagent-engineering-toolkit --skill karpathy-guidelines",{"tool":37,"command":38},"bunx","bunx skills add dexter-cnx\u002Fagent-engineering-toolkit --skill karpathy-guidelines",{"tool":40,"command":41},"pnpm","pnpm dlx skills add dexter-cnx\u002Fagent-engineering-toolkit --skill karpathy-guidelines","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fdexter-cnx\u002Fagent-engineering-toolkit\u002Fblob\u002Fmain\u002Foverlays\u002Fagent-karpathy\u002Fskills\u002Fkarpathy-guidelines\u002FSKILL.md",true,[],[46,73,93,102,114,126],{"skillNo":47,"name":48,"nameZh":6,"description":49,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":50,"categories":51,"keywords":52,"publishedAt":20},"01M6P4P9S9017DTSFMRKGA7XW2","toolkit-i18n-operator","Use the toolkit-i18n CLI for repeated Flutter localization workflows with compact outputs, CSV validation, key coverage reporting, generated JSON exports, and safe non-mutating behavior.","https:\u002F\u002Fgithub.com\u002Fdexter-cnx\u002Fagent-engineering-toolkit\u002Fblob\u002Fmain\u002Ftools\u002Ftoolkit-i18n\u002Fskills\u002Ftoolkit-i18n-operator\u002FSKILL.md",[],[53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72],"toolkit i18n operator","i18n","operator","toolkit","flutter localization","cli","safe","workflows","for","validation","with","csv","compact","reporting","json","flutter","behavior","coverage","generated","use",{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":89,"publishedAt":20},"01M6P4NREKE3P0XYV3RBBWC987","toolkit-design-operator","Use the toolkit-design CLI for repeated repository workflows.","https:\u002F\u002Fgithub.com\u002Fdexter-cnx\u002Fagent-engineering-toolkit\u002Fblob\u002Fmain\u002Ftools\u002Ftoolkit-design\u002Fskills\u002Ftoolkit-design-operator\u002FSKILL.md",[79,84],{"categoryNo":80,"categoryName":81,"taxonomyType":82,"taxonomyTypeDesc":83},"01JBSEEDCAT000000000000015","视觉设计","SCENARIO","使用场景",{"categoryNo":85,"categoryName":86,"taxonomyType":87,"taxonomyTypeDesc":88},"01JBSEEDCAT000000000000032","设计师","PROFESSION","适用职业",[90,55,56,91,61,58,60,92,72],"toolkit design operator","design","repository",{"skillNo":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"publishedAt":20},"01M6P4NNSQ09QFPYG2PMPWEA94","toolkit-ci-operator","Use the toolkit-ci CLI for repeated repository workflows.","https:\u002F\u002Fgithub.com\u002Fdexter-cnx\u002Fagent-engineering-toolkit\u002Fblob\u002Fmain\u002Ftools\u002Ftoolkit-ci\u002Fskills\u002Ftoolkit-ci-operator\u002FSKILL.md",[],[100,101,55,56,92,58,60,72,61],"toolkit ci operator","ci",{"skillNo":103,"name":104,"nameZh":6,"description":105,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":106,"categories":107,"keywords":108,"publishedAt":20},"01M6P4NK636991P6WWQ6X8HDYX","toolkit-arch-operator","Use the toolkit-arch CLI for repeated repository workflows, including Flutter i18n usage scans, layer checks, and localization coverage.","https:\u002F\u002Fgithub.com\u002Fdexter-cnx\u002Fagent-engineering-toolkit\u002Fblob\u002Fmain\u002Ftools\u002Ftoolkit-arch\u002Fskills\u002Ftoolkit-arch-operator\u002FSKILL.md",[],[109,110,56,55,54,111,112,61,72,70,68,113,92,60,58],"toolkit arch operator","arch","usage","layer","localization",{"skillNo":115,"name":116,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":117,"categories":118,"keywords":122,"publishedAt":20},"01M6P4NF18XGVSG2NR01E1DX0R","web-testing","https:\u002F\u002Fgithub.com\u002Fdexter-cnx\u002Fagent-engineering-toolkit\u002Fblob\u002Fmain\u002Foverlays\u002Fweb-frontend\u002Fskills\u002Fweb-testing\u002Fskill.md",[119],{"categoryNo":120,"categoryName":121,"taxonomyType":82,"taxonomyTypeDesc":83},"01JBSEEDCAT000000000000008","测试与质量",[123,124,125],"web testing","web","testing",{"skillNo":127,"name":128,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":6,"repositoryNo":11,"starsCount":12,"licenseSpdxId":13,"licenseFileUrl":6,"sourceHtmlUrl":129,"categories":130,"keywords":131,"publishedAt":20},"01M6P4MVK07EJ75FQTFN54AGPM","web-state","https:\u002F\u002Fgithub.com\u002Fdexter-cnx\u002Fagent-engineering-toolkit\u002Fblob\u002Fmain\u002Foverlays\u002Fweb-frontend\u002Fskills\u002Fweb-state\u002Fskill.md",[],[124,132,133],"state","web state"]