[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78FFYSHYB5AD6SVC7PER9ED":3,"skill-same:01M78FFYSHYB5AD6SVC7PER9ED":44},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M78FFYSHYB5AD6SVC7PER9ED","honest-artifacts",null,"Make outputs honest and re-derivable. Use when committing constants, thresholds, or tuned parameters; when reporting a number or a benchmark result; when a result came from a manual one-off process; or when optimizing against a metric. Also use when tempted to present an estimate as if it were verified.","songjiun10-collab\u002FHncs","songjiun10-collab","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F293141730?v=4","01M78FEF4NYSK5BRK9HBBT64X1","01M78FEF5AKPHSAAMYT6249TR7",16,"MIT","https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab\u002FHncs\u002Fblob\u002Fmain\u002Fsenior-engineer-mindset\u002Fhonest-artifacts\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: honest-artifacts\ndescription: Make outputs honest and re-derivable. Use when committing constants, thresholds, or tuned parameters; when reporting a number or a benchmark result; when a result came from a manual one-off process; or when optimizing against a metric. Also use when tempted to present an estimate as if it were verified.\n---\n\n# Honest Artifacts\n\n## Separate what you know from what you estimated\n\nThe most dangerous code isn't wrong code — it's code where **you can't tell how much of it is verified.** Put a validated value next to a guessed one with no distinction, and the next person either trusts both or doubts both.\n\n- Does this constant, parameter, or default have a real basis, or is it \"this is probably close enough\"?\n- If there's no basis, is that fact **explicitly recorded** in code or docs? (`# UNVERIFIED: pulled from X, not independently checked`)\n- Are you presenting an estimate as if it were certain?\n- A number that gets quoted outside this task — in a design doc, a status update, another team's decision — carries your confidence level with it whether you state it or not. Label unverified numbers before they travel, not after someone has already built a decision on top of them.\n- **Distinguished\u002FFellow angle:** A number or methodology solid enough to appear in a conference talk, a blog post, or an open-source release needs to survive scrutiny from people with zero institutional memory of how it was produced — label what's unverified before it becomes the industry's reference point, not after.\n- **Executive angle (CTO\u002FVP-Eng):** A number that ends up in a customer SLA, a sales claim, or a board deck carries legal and reputational risk if it can't be reproduced on demand — verify it before it leaves engineering, not after a customer or investor has already been quoted it.\n- Record the attempts that didn't pan out too — the next person won't hit the same wall again. **Failure is data, not something to hide.**\n\n## Reproducibility\n\n\"It works\" and \"it can be rebuilt\" are different claims. A result from a one-off manual process is a black box the moment that person is gone.\n\n- Is there a procedure to rebuild this result (a number, a dataset, an artifact) from scratch?\n- Does that procedure run as **one command**, or does it live only in someone's head?\n- Will the number you're committing now reproduce from the same inputs six months from now? (random seeds, external data drift, version differences)\n\n## Hncs's own version of this practice\n\nThis isn't hypothetical here. Each `apply_*` file's docstring carries\nits full fitting history (sample size, sources, what was tried and\nrejected) — new data gets a dated paragraph appended, the old one is\nnever rewritten. `hybrid_engine\u002FEVALUATION.md` records every experiment,\nwin\u002Floss\u002Finconclusive, on purpose: \"the failures are the asset,\" ~20\nrejected approaches kept on the record so the next session doesn't\nre-run them. When a finding later turns out wrong, the fix is a dated\ncorrection blockquote added **in place** — original claim, correction\ndate, and how the mistake was caught, right next to the thing it\ncorrects — so the original claim stays visible as history instead of\nbeing edited away. That's what \"record the attempts that\ndidn't pan out\" looks like as a working convention instead of an\naspiration.\n\n## Metric traps\n\nOptimize a score long enough and the score improves while the actual goal drifts away.\n\n- Did this change make the **thing you're measuring** better, or did it just fit **the way you're measuring it**? (overfitting)\n- Is the sample size big enough? If not, a **conservative choice** beats a complex one with a marginally better score\n- Did anything get worse while this metric improved? (a hidden tradeoff — speed up, accuracy down)\n- **\"It got better\" is not a result.** Give the number.\n","# Honest Artifacts\n\n## Separate what you know from what you estimated\n\nThe most dangerous code isn't wrong code — it's code where **you can't tell how much of it is verified.** Put a validated value next to a guessed one with no distinction, and the next person either trusts both or doubts both.\n\n- Does this constant, parameter, or default have a real basis, or is it \"this is probably close enough\"?\n- If there's no basis, is that fact **explicitly recorded** in code or docs? (`# UNVERIFIED: pulled from X, not independently checked`)\n- Are you presenting an estimate as if it were certain?\n- A number that gets quoted outside this task — in a design doc, a status update, another team's decision — carries your confidence level with it whether you state it or not. Label unverified numbers before they travel, not after someone has already built a decision on top of them.\n- **Distinguished\u002FFellow angle:** A number or methodology solid enough to appear in a conference talk, a blog post, or an open-source release needs to survive scrutiny from people with zero institutional memory of how it was produced — label what's unverified before it becomes the industry's reference point, not after.\n- **Executive angle (CTO\u002FVP-Eng):** A number that ends up in a customer SLA, a sales claim, or a board deck carries legal and reputational risk if it can't be reproduced on demand — verify it before it leaves engineering, not after a customer or investor has already been quoted it.\n- Record the attempts that didn't pan out too — the next person won't hit the same wall again. **Failure is data, not something to hide.**\n\n## Reproducibility\n\n\"It works\" and \"it can be rebuilt\" are different claims. A result from a one-off manual process is a black box the moment that person is gone.\n\n- Is there a procedure to rebuild this result (a number, a dataset, an artifact) from scratch?\n- Does that procedure run as **one command**, or does it live only in someone's head?\n- Will the number you're committing now reproduce from the same inputs six months from now? (random seeds, external data drift, version differences)\n\n## Hncs's own version of this practice\n\nThis isn't hypothetical here. Each `apply_*` file's docstring carries\nits full fitting history (sample size, sources, what was tried and\nrejected) — new data gets a dated paragraph appended, the old one is\nnever rewritten. `hybrid_engine\u002FEVALUATION.md` records every experiment,\nwin\u002Floss\u002Finconclusive, on purpose: \"the failures are the asset,\" ~20\nrejected approaches kept on the record so the next session doesn't\nre-run them. When a finding later turns out wrong, the fix is a dated\ncorrection blockquote added **in place** — original claim, correction\ndate, and how the mistake was caught, right next to the thing it\ncorrects — so the original claim stays visible as history instead of\nbeing edited away. That's what \"record the attempts that\ndidn't pan out\" looks like as a working convention instead of an\naspiration.\n\n## Metric traps\n\nOptimize a score long enough and the score improves while the actual goal drifts away.\n\n- Did this change make the **thing you're measuring** better, or did it just fit **the way you're measuring it**? (overfitting)\n- Is the sample size big enough? If not, a **conservative choice** beats a complex one with a marginally better score\n- Did anything get worse while this metric improved? (a hidden tradeoff — speed up, accuracy down)\n- **\"It got better\" is not a result.** Give the number.\n",{"name":5,"description":7},"main","513567c2b0eb44c159ebac74b0491bdcf823c377","senior-engineer-mindset\u002Fhonest-artifacts",2,"https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab\u002FHncs","https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab","2026-11-07 16:34:56",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add songjiun10-collab\u002FHncs --skill honest-artifacts",{"tool":36,"command":37},"bunx","bunx skills add songjiun10-collab\u002FHncs --skill honest-artifacts",{"tool":39,"command":40},"pnpm","pnpm dlx skills add songjiun10-collab\u002FHncs --skill honest-artifacts","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab\u002FHncs\u002Fblob\u002Fmain\u002Fsenior-engineer-mindset\u002Fhonest-artifacts\u002FSKILL.md",true,[],[45,52,59,66,73,80],{"skillNo":46,"name":47,"nameZh":6,"description":48,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":49,"categories":50,"keywords":51,"publishedAt":18},"01M78FJ7DGE8S2MHJQY4RVTPMF","widen-the-solution-space","Deliberately widen the set of candidate approaches before committing to one. Use before irreversible decisions (architecture, DB schema, data format, public API), when the first solution feels complicated or awkward, when the user is deciding rather than instructing (\"what's the best way to do this?\"), or when a previous attempt at the same problem failed. Skip for routine work with one obvious answer.","https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab\u002FHncs\u002Fblob\u002Fmain\u002Fsenior-engineer-mindset\u002Fwiden-the-solution-space\u002FSKILL.md",[],[],{"skillNo":53,"name":54,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":56,"categories":57,"keywords":58,"publishedAt":18},"01M78FJ2D6HMAGS8E11VBD4NS0","weigh-tradeoffs","Compare real alternatives explicitly and record why one was chosen. Use when picking between libraries, data structures, architectures, or storage options; when a decision is hard to reverse (schema, public API, data format, dependency); or when the user asks \"which is better, A or B\". Also use to decide how much deliberation a decision even deserves.","https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab\u002FHncs\u002Fblob\u002Fmain\u002Fsenior-engineer-mindset\u002Fweigh-tradeoffs\u002FSKILL.md",[],[],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":18},"01M78FHX1P3QMTMC8WPQS0Z669","verify-before-claiming","Run the verification and read its output before saying anything is done, fixed, working, or passing. Use before any completion claim, before committing or opening a PR, before moving to the next task, and after a subagent or tool reports success. Triggers on the words \"should work\", \"fixed\", \"done\", \"passing\", \"looks good\" — and on the urge to celebrate.","https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab\u002FHncs\u002Fblob\u002Fmain\u002Fsenior-engineer-mindset\u002Fverify-before-claiming\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":18},"01M78FHSBQ30GH1EMSBA8NJZNX","verifiability-first","Turn the task into something verifiable and declare the success criterion before seeing results. Use when the request is vague about \"done\" (\"make it work\", \"improve this\"), when writing tests, when fixing a bug (write the failing reproduction first), or when comparing two approaches. Critical before any measurement or benchmark, where deciding the bar after seeing the numbers guarantees a \"success\".","https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab\u002FHncs\u002Fblob\u002Fmain\u002Fsenior-engineer-mindset\u002Fverifiability-first\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":18},"01M78FHN4P77KF3T47M1NBT7XA","threat-and-scale-check","Check trust boundaries, scale behavior, and layered safety. Use when handling user input, external API responses, files, or auth; when designing permissions, validation, or any safety mechanism; when data volume could grow; or when writing loops that touch I\u002FO. Also use when reviewing whether a single safeguard is the only thing standing between a mistake and a disaster.","https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab\u002FHncs\u002Fblob\u002Fmain\u002Fsenior-engineer-mindset\u002Fthreat-and-scale-check\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":18},"01M78FHGGZKPTBRVWJK83X18CY","surgical-change","Keep every changed line traceable to the request. Use when editing existing code, especially code you didn't write; when tempted to clean up something you noticed while reading; when a diff is growing beyond the ask; or when deciding whether to delete code that looks dead.","https:\u002F\u002Fgithub.com\u002Fsongjiun10-collab\u002FHncs\u002Fblob\u002Fmain\u002Fsenior-engineer-mindset\u002Fsurgical-change\u002FSKILL.md",[],[]]