[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7DEEYP118HYVHS8BJSGAQHB":3,"skill-same:01M7DEEYP118HYVHS8BJSGAQHB":70},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":28,"ownerHtmlUrl":29,"ownerBlogUrl":28,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M7DEEYP118HYVHS8BJSGAQHB","szl-experiment-replay",null,"Replays a pinned, bounded CSV column mean against a frozen reference result and writes an explicit match, divergence or incomplete receipt. Use when a researcher needs to rerun a small tabular result offline and distinguish numeric disagreement from changed evidence.","szl-holdings\u002Fszl-skills","szl-holdings","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F273360816?v=4","01M7DEC83S24HX3V8G1JSWKHD0","SZL Holdings","01M7DEC85084EGHEJH2KMAA53K",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-experiment-replay\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: szl-experiment-replay\ndescription: \"Replays a pinned, bounded CSV column mean against a frozen reference result and writes an explicit match, divergence or incomplete receipt. Use when a researcher needs to rerun a small tabular result offline and distinguish numeric disagreement from changed evidence.\"\nlicense: Apache-2.0\n---\n\n# Experiment replay\n\nUse Python 3.10+ and the standard library to replay one fixed operation on local CSV data. This\nexecutes the skill's built-in `csv_column_mean_v1`; it does not execute a supplied script,\nshell command, notebook, plugin or arbitrary Python. It uses no network or credentials.\n\n## Run\n\nFrom this skill directory, use a writable copy of the synthetic assets or a study root of your own:\n\n```bash\npython -I -B scripts\u002Frun.py prepare assets\u002Fdeclaration.json --root . --output replay-pin.json\npython -I -B scripts\u002Frun.py replay replay-pin.json --root . --receipt replay-receipt.json\n```\n\nEach output path must be new and relative to `--root`; the runner refuses overwrite. The\ndeclaration names a CSV input, JSON reference result, column and absolute decimal tolerance.\n`prepare` freezes the exact input, reference and built-in `scripts\u002Fengine.py` hashes. The CLI\nloads that engine before `replay` compares its hash, then checks the input and reference hashes\nbefore calculating the mean. This detects drift after loading a trusted package; it is not a\npre-execution integrity, authenticity or sandbox gate. `MATCH` means the observed mean\nfell within the declared tolerance; `DIVERGED` means it did not. Changed or missing files yield\n`INCOMPLETE`; invalid pins yield `REFUSED`. Exit codes: 0 match, 1 divergence, 2 incomplete or\nrefused. With a valid root and a safe, new receipt path, a missing, unreadable or malformed\non-disk pin also yields a retained `REFUSED` receipt. An invalid pin is not identified by its\nunvalidated claimed digest: the receipt's `pin_sha256` is null. If the root or receipt path is\nunsafe or unwritable, the CLI exits 2 but cannot retain a receipt. A reference with a different\nvalue must be deliberately prepared as a new pin to test divergence; editing a previously pinned\nreference yields `INCOMPLETE`.\n\nRetain the pin, receipt and input\u002Freference bytes together. The self-digests catch accidental\nchanges but are not signatures. Receipts explicitly say `SAME_HOST_LOCAL`, `signed: false`,\n`independent_witness: false` and `scientific_claims_verified: false`. The runner does not prove\nthe experiment design, data collection, author identity or another machine's reproduction.\n\n## Bounds and relationship to other skills\n\nInput CSV: at most 256 KiB, 10,000 rows, 64 columns; reference JSON: at most 4 KiB. Paths are\ncanonical relative paths under the selected root. Symlinks, junctions, other reparse points,\nspecial files, traversal, credential-like names and existing output files are refused. This is\nfor a cooperative local filesystem, not an adversarial race-safe sandbox.\n\nThis is a CLI-only skill; `scripts\u002Fengine.py` is not a Claude Science sidecar.\nSource tests and an SDK test double do not verify that the real Claude Science host can run\nthe CLI or attach this skill. Keep host registration and task evaluation `NOT_VERIFIED` until\nthat control plane supplies a gate, readback and pilot evidence.\n\nUse `szl-reproducibility-capsule` to inventory broader source, environment and analysis plan\nfiles. Its replay declaration remains inert; this skill independently executes only the fixed\nCSV mean operation. For publication, arrange independent retention and a separate witness.\n","# Experiment replay\n\nUse Python 3.10+ and the standard library to replay one fixed operation on local CSV data. This\nexecutes the skill's built-in `csv_column_mean_v1`; it does not execute a supplied script,\nshell command, notebook, plugin or arbitrary Python. It uses no network or credentials.\n\n## Run\n\nFrom this skill directory, use a writable copy of the synthetic assets or a study root of your own:\n\n```bash\npython -I -B scripts\u002Frun.py prepare assets\u002Fdeclaration.json --root . --output replay-pin.json\npython -I -B scripts\u002Frun.py replay replay-pin.json --root . --receipt replay-receipt.json\n```\n\nEach output path must be new and relative to `--root`; the runner refuses overwrite. The\ndeclaration names a CSV input, JSON reference result, column and absolute decimal tolerance.\n`prepare` freezes the exact input, reference and built-in `scripts\u002Fengine.py` hashes. The CLI\nloads that engine before `replay` compares its hash, then checks the input and reference hashes\nbefore calculating the mean. This detects drift after loading a trusted package; it is not a\npre-execution integrity, authenticity or sandbox gate. `MATCH` means the observed mean\nfell within the declared tolerance; `DIVERGED` means it did not. Changed or missing files yield\n`INCOMPLETE`; invalid pins yield `REFUSED`. Exit codes: 0 match, 1 divergence, 2 incomplete or\nrefused. With a valid root and a safe, new receipt path, a missing, unreadable or malformed\non-disk pin also yields a retained `REFUSED` receipt. An invalid pin is not identified by its\nunvalidated claimed digest: the receipt's `pin_sha256` is null. If the root or receipt path is\nunsafe or unwritable, the CLI exits 2 but cannot retain a receipt. A reference with a different\nvalue must be deliberately prepared as a new pin to test divergence; editing a previously pinned\nreference yields `INCOMPLETE`.\n\nRetain the pin, receipt and input\u002Freference bytes together. The self-digests catch accidental\nchanges but are not signatures. Receipts explicitly say `SAME_HOST_LOCAL`, `signed: false`,\n`independent_witness: false` and `scientific_claims_verified: false`. The runner does not prove\nthe experiment design, data collection, author identity or another machine's reproduction.\n\n## Bounds and relationship to other skills\n\nInput CSV: at most 256 KiB, 10,000 rows, 64 columns; reference JSON: at most 4 KiB. Paths are\ncanonical relative paths under the selected root. Symlinks, junctions, other reparse points,\nspecial files, traversal, credential-like names and existing output files are refused. This is\nfor a cooperative local filesystem, not an adversarial race-safe sandbox.\n\nThis is a CLI-only skill; `scripts\u002Fengine.py` is not a Claude Science sidecar.\nSource tests and an SDK test double do not verify that the real Claude Science host can run\nthe CLI or attach this skill. Keep host registration and task evaluation `NOT_VERIFIED` until\nthat control plane supplies a gate, readback and pilot evidence.\n\nUse `szl-reproducibility-capsule` to inventory broader source, environment and analysis plan\nfiles. Its replay declaration remains inert; this skill independently executes only the fixed\nCSV mean operation. For publication, arrange independent retention and a separate witness.\n",{"name":5,"license":15,"description":7},"main","8dd614e1f5097564340e36915f849d7d893cd68a","skills\u002Fszl-experiment-replay","https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills","https:\u002F\u002Fa-11-oy.com","https:\u002F\u002Fgithub.com\u002Fszl-holdings","Governed AI infrastructure for inspectable decisions: source, runtime state, receipts, and proof boundaries.","2026-11-15 03:52:34",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add szl-holdings\u002Fszl-skills --skill szl-experiment-replay",{"tool":38,"command":39},"bunx","bunx skills add szl-holdings\u002Fszl-skills --skill szl-experiment-replay",{"tool":41,"command":42},"pnpm","pnpm dlx skills add szl-holdings\u002Fszl-skills --skill szl-experiment-replay","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-experiment-replay\u002FSKILL.md",true,[46,52,56,60,66],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"assets\u002Fdeclaration.json","ASSET","资产文件",207,"https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-experiment-replay\u002Fassets\u002Fdeclaration.json",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"assets\u002Fexpected.json",62,"https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-experiment-replay\u002Fassets\u002Fexpected.json",{"path":57,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":58,"sourceHtmlUrl":59},"assets\u002Fsynthetic.csv",51,"https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-experiment-replay\u002Fassets\u002Fsynthetic.csv",{"path":61,"resourceType":62,"resourceTypeDesc":63,"sizeBytes":64,"sourceHtmlUrl":65},"scripts\u002Fengine.py","SCRIPT","脚本",13695,"https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-experiment-replay\u002Fscripts\u002Fengine.py",{"path":67,"resourceType":62,"resourceTypeDesc":63,"sizeBytes":68,"sourceHtmlUrl":69},"scripts\u002Frun.py",2856,"https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-experiment-replay\u002Fscripts\u002Frun.py",[71,78,85,92,98,103],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":19},"01M7DENG22R761RTDSG11WWJB3","szl-unit-invariants","Checks declared SI dimensions, allowlisted unit conversions, expected ranges and fixed pointwise invariants (equal, sum, difference, product, ratio) over selected measurement or simulation columns, flagging incompatible quantities, scale errors of 1000x and broken conservation checks. Use when merging instrument exports with different units, validating simulation output against conservation laws, or when a reviewer asks how unit errors were ruled out. Not an uncertainty-propagation or symbolic tool.","https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-unit-invariants\u002FSKILL.md",[],[],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":19},"01M7DENAVBJ6GSCRT3E1PWK56D","szl-uncertainty-lineage","Propagate declared measurement uncertainty through a small arithmetic dependency graph using a first-order delta approximation. Use when a researcher needs to see how corrected inputs or shared-source correlation change a reported standard uncertainty; require unit checks separately.","https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-uncertainty-lineage\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":19},"01M7DEN2M7P59RTM6RJE99PGN8","szl-typesafe-ai","TypeSafe\u002FJev: Choice\u002FNoul\u002FScore in code, not TypeScript types. System One models turn natural language and application state into probabilities that software can compose. Use when building routing, ranking, extraction, verification, or replacing an LLM prompt-and-parse step with structured decisions; also when brainstorming what TypeSafe could make possible. Do not use for TypeScript\u002FJava\u002FRust static types, Zod\u002FPydantic\u002Fmypy\u002FJSON Schema\u002Fprotobuf codegen, or asking whether a pipeline is type-safe. At SZL, Jev is an optional second reader only.","https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-typesafe-ai\u002FSKILL.md",[],[],{"skillNo":93,"name":94,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":19},"01M7DEN1C9G4H12N9HRHYK2HZ7","demo","https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-skill-update-review\u002Fassets\u002Fexample\u002Fold-package\u002Fskills\u002Fdemo\u002FSKILL.md",[],[],{"skillNo":99,"name":94,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":19},"01M7DEMZYRXT5CG2XPXERAM218","https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-skill-update-review\u002Fassets\u002Fexample\u002Fnew-package\u002Fskills\u002Fdemo\u002FSKILL.md",[],[],{"skillNo":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":19},"01M7DEMVH2NMG4PS0DZTMGATAF","szl-skill-update-review","Compare pinned local skill packages using inventories and a retained lock; report changed skills, files, declarations and reruns without approving safety or validity.","https:\u002F\u002Fgithub.com\u002Fszl-holdings\u002Fszl-skills\u002Fblob\u002Fmain\u002Fskills\u002Fszl-skill-update-review\u002FSKILL.md",[],[]]