[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7A00HCWZKYMXD4TYCXDPFA4":3,"skill-same:01M7A00HCWZKYMXD4TYCXDPFA4":65},{"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":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":6,"ownerDescription":31,"updatedAt":32,"install":33,"resources":46},"01M7A00HCWZKYMXD4TYCXDPFA4","ageval-config-package",null,"Author ageval datasets (ageval.yaml + tasks\u002F*\u002Ftask.yaml, run.py\u002Fevaluator.py, profiles.yaml, environment kinds local|docker|e2b|ssh|daytona, limits, gold isolation, script vs LLM-as-judge scoring). Dataset README Run commands start with `ageval` (CLI already installed). Never `uv run ageval` or `.venv\u002Fbin\u002Fageval`. Triggers: ageval.yaml, task.yaml, profiles.yaml, dataset, environment kind, evaluator.py. Never secrets in yaml. Never provider.kind.","ZJU-REAL\u002Fageval","ZJU-REAL","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F203928864?v=4","01M7A005CH0Y4HRGXVRH1G511A","REAL Lab","01M7A005D4AE6Y2WT2ZJQDVKVM",81,"MIT","https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fageval-config-package\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: ageval-config-package\ndescription: >\n  Author ageval datasets (ageval.yaml + tasks\u002F*\u002Ftask.yaml, run.py\u002Fevaluator.py,\n  profiles.yaml, environment kinds local|docker|e2b|ssh|daytona, limits, gold isolation,\n  script vs LLM-as-judge scoring). Dataset README Run commands start with `ageval`\n  (CLI already installed). Never `uv run ageval` or `.venv\u002Fbin\u002Fageval`. Triggers:\n  ageval.yaml, task.yaml, profiles.yaml, dataset, environment kind, evaluator.py.\n  Never secrets in yaml. Never provider.kind.\n---\n\n# Config \u002F dataset + task\n\nConfig Core is the only normative reader. Delivery unit is a **dataset** root.\n\n```text\nmy-dataset\u002F                 # CLI path (ageval.dataset\u002F1)\n├── ageval.yaml\n├── profiles.yaml           # THE job (lock\u002Frun default). ageval publish also packs every root profiles*.yaml\n├── env.example\n├── shared\u002Flib\u002F             # optional; import shared.lib.*\n└── tasks\u002F\u003Ctask_id>\u002F\n    ├── task.yaml           # ageval.task\u002F1 — roles + intent\n    ├── run.py              # async def run(ctx)\n    ├── evaluator.py\n    ├── environment\u002F        # Dockerfile, setup.sh, compose\n    ├── evaluation\u002F         # gold — not for Agent\n    └── data\u002F               # seed\n```\n\nCLI: `ageval lock|run \u003Cdataset-root> --task \u003Cid>`.\n\n**Reader-facing README \u002F docs:** commands start with `ageval`. The reader already has the CLI. Do not write `uv run ageval`, `uv run ageval lock`, `.venv\u002Fbin\u002Fageval`, or `AGEVAL=\u002Fpath\u002Fto\u002F...` wrappers. Agent-internal notes (this skill, AGENTS.md) may name a venv binary; dataset README may not.\n\nDataset README introduces **this dataset** (identity, `profiles.yaml`, layout, `ageval lock|run`, conversion, known gaps). It does **not** document local `.ageval\u002Fsuite-runs` ids, P\u002FF\u002FE counts, or leaderboard rows. Those are `upload-suite` later, not package copy.\n\nFiles present are recognized. Do not re-declare `run.py` entrypoints in yaml. Do not put `executor` \u002F `api_key` on the member task. Do not write `provider.kind` or `assurance`.\n\n```yaml\n# ageval.yaml\nformat: ageval.dataset\u002F1\ndataset_id: org\u002Fmy-suite\nversion: \"0.1.0\"\ntasks:\n  root: tasks\n```\n\n```yaml\n# profiles.yaml\nformat: ageval.profiles\u002F1\nenvironment: docker\nagent_profiles:\n  solver:\n    executor: acp\n    api_key: ${ZHIPU_API_KEY}\n    options: { entry: pi }\n    extensions:\n      - plugin: acp\n      - plugin: docker\n```\n\nGold lives in `evaluation\u002F` and is uploaded at evaluate, not before. `setup.sh` is the last environment slot.\n\nTwo scoring styles in `evaluator.py` (same barrier; PASS still the return value):\n\n| Style | When | Evidence |\n| --- | --- | --- |\n| Deterministic script | Default. Compare artifacts to gold. No `Agent.session`. | `result.json`. No `evaluation\u002Fobservation.jsonl`. Optional `checks` via `evaluation_check(...)` → `evaluation\u002Fchecks.json`. |\n| LLM-as-judge | Extra role on the **same** `profiles.yaml` (e.g. `judge`). After gold lands, `Agent.session(\"judge\").invoke`. | `evaluation\u002Fobservation.jsonl` (no `user` rows). Not merged into Agent `trajectory.jsonl`. May also return `checks`. |\n\nJudge binding stays in `profiles.yaml` (not on the member task). `limits.agent_invocations` counts run-phase invokes only; judge invokes are bounded by `limits.evaluate_seconds`. `limits.agent_concurrency` caps simultaneous run-phase executor invokes (default 1). Judge admission stays serial. SDK session into the evaluator process is projected on `environment: local`. No `Agent.session` = script path.\n\n`--profiles` swaps the job for **this** lock\u002Frun. `--agent` is mutually exclusive with `--profiles`. `--model` requires `--agent`.\n\n### `profiles.yaml` vs extra `profiles*.yaml`\n\n| | Default job (`ageval run \u003Croot>`) | `ageval publish` \u002F packageDigest |\n| --- | --- | --- |\n| `profiles.yaml` | **this file only** | included |\n| sibling `profiles*.yaml` at dataset root | ignored unless `--profiles` | **included** (`member_paths_for_digest` \u002F tarball). Overlays those docs declare also enter. |\n\nHub consumers who `ageval run official\u002F…@ver` without `--profiles` get **`profiles.yaml`**. Extra `profiles.acp.*.yaml` at the root still **ship in the package** and change digest. Experiment jobs that must not be on Hub do not live as `profiles*.yaml` under the dataset root.\n\nHub comparability still uses suite fingerprint \u002F `upload-suite`. Do not add `n_attempts` to yaml.\n\nField catalog: [references\u002Fageval-yaml.md](references\u002Fageval-yaml.md). Isolation: [references\u002Fisolation.md](references\u002Fisolation.md). Conversion: [references\u002Fconversion.md](references\u002Fconversion.md).\n","# Config \u002F dataset + task\n\nConfig Core is the only normative reader. Delivery unit is a **dataset** root.\n\n```text\nmy-dataset\u002F                 # CLI path (ageval.dataset\u002F1)\n├── ageval.yaml\n├── profiles.yaml           # THE job (lock\u002Frun default). ageval publish also packs every root profiles*.yaml\n├── env.example\n├── shared\u002Flib\u002F             # optional; import shared.lib.*\n└── tasks\u002F\u003Ctask_id>\u002F\n    ├── task.yaml           # ageval.task\u002F1 — roles + intent\n    ├── run.py              # async def run(ctx)\n    ├── evaluator.py\n    ├── environment\u002F        # Dockerfile, setup.sh, compose\n    ├── evaluation\u002F         # gold — not for Agent\n    └── data\u002F               # seed\n```\n\nCLI: `ageval lock|run \u003Cdataset-root> --task \u003Cid>`.\n\n**Reader-facing README \u002F docs:** commands start with `ageval`. The reader already has the CLI. Do not write `uv run ageval`, `uv run ageval lock`, `.venv\u002Fbin\u002Fageval`, or `AGEVAL=\u002Fpath\u002Fto\u002F...` wrappers. Agent-internal notes (this skill, AGENTS.md) may name a venv binary; dataset README may not.\n\nDataset README introduces **this dataset** (identity, `profiles.yaml`, layout, `ageval lock|run`, conversion, known gaps). It does **not** document local `.ageval\u002Fsuite-runs` ids, P\u002FF\u002FE counts, or leaderboard rows. Those are `upload-suite` later, not package copy.\n\nFiles present are recognized. Do not re-declare `run.py` entrypoints in yaml. Do not put `executor` \u002F `api_key` on the member task. Do not write `provider.kind` or `assurance`.\n\n```yaml\n# ageval.yaml\nformat: ageval.dataset\u002F1\ndataset_id: org\u002Fmy-suite\nversion: \"0.1.0\"\ntasks:\n  root: tasks\n```\n\n```yaml\n# profiles.yaml\nformat: ageval.profiles\u002F1\nenvironment: docker\nagent_profiles:\n  solver:\n    executor: acp\n    api_key: ${ZHIPU_API_KEY}\n    options: { entry: pi }\n    extensions:\n      - plugin: acp\n      - plugin: docker\n```\n\nGold lives in `evaluation\u002F` and is uploaded at evaluate, not before. `setup.sh` is the last environment slot.\n\nTwo scoring styles in `evaluator.py` (same barrier; PASS still the return value):\n\n| Style | When | Evidence |\n| --- | --- | --- |\n| Deterministic script | Default. Compare artifacts to gold. No `Agent.session`. | `result.json`. No `evaluation\u002Fobservation.jsonl`. Optional `checks` via `evaluation_check(...)` → `evaluation\u002Fchecks.json`. |\n| LLM-as-judge | Extra role on the **same** `profiles.yaml` (e.g. `judge`). After gold lands, `Agent.session(\"judge\").invoke`. | `evaluation\u002Fobservation.jsonl` (no `user` rows). Not merged into Agent `trajectory.jsonl`. May also return `checks`. |\n\nJudge binding stays in `profiles.yaml` (not on the member task). `limits.agent_invocations` counts run-phase invokes only; judge invokes are bounded by `limits.evaluate_seconds`. `limits.agent_concurrency` caps simultaneous run-phase executor invokes (default 1). Judge admission stays serial. SDK session into the evaluator process is projected on `environment: local`. No `Agent.session` = script path.\n\n`--profiles` swaps the job for **this** lock\u002Frun. `--agent` is mutually exclusive with `--profiles`. `--model` requires `--agent`.\n\n### `profiles.yaml` vs extra `profiles*.yaml`\n\n| | Default job (`ageval run \u003Croot>`) | `ageval publish` \u002F packageDigest |\n| --- | --- | --- |\n| `profiles.yaml` | **this file only** | included |\n| sibling `profiles*.yaml` at dataset root | ignored unless `--profiles` | **included** (`member_paths_for_digest` \u002F tarball). Overlays those docs declare also enter. |\n\nHub consumers who `ageval run official\u002F…@ver` without `--profiles` get **`profiles.yaml`**. Extra `profiles.acp.*.yaml` at the root still **ship in the package** and change digest. Experiment jobs that must not be on Hub do not live as `profiles*.yaml` under the dataset root.\n\nHub comparability still uses suite fingerprint \u002F `upload-suite`. Do not add `n_attempts` to yaml.\n\nField catalog: [references\u002Fageval-yaml.md](references\u002Fageval-yaml.md). Isolation: [references\u002Fisolation.md](references\u002Fisolation.md). Conversion: [references\u002Fconversion.md](references\u002Fconversion.md).\n",{"name":5,"description":7},"main","ccda2a7af15ac19fde705e20d3b267c645330d4a","skills\u002Fageval-config-package",3,"https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval","https:\u002F\u002Fzju-real.github.io\u002Fageval\u002F","https:\u002F\u002Fgithub.com\u002FZJU-REAL","REAL Lab is a research team at Zhejiang University focusing on Reasoning, Embodied, Agentic and Lifelong Evolving AI.","2026-11-14 01:27:17",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add ZJU-REAL\u002Fageval --skill ageval-config-package",{"tool":39,"command":40},"bunx","bunx skills add ZJU-REAL\u002Fageval --skill ageval-config-package",{"tool":42,"command":43},"pnpm","pnpm dlx skills add ZJU-REAL\u002Fageval --skill ageval-config-package","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fageval-config-package\u002FSKILL.md",true,[47,53,57,61],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"agents\u002Fopenai.yaml","REFERENCE","参考文档",246,"https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fageval-config-package\u002Fagents\u002Fopenai.yaml",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"references\u002Fageval-yaml.md",1730,"https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fageval-config-package\u002Freferences\u002Fageval-yaml.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"references\u002Fconversion.md",3648,"https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fageval-config-package\u002Freferences\u002Fconversion.md",{"path":62,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":63,"sourceHtmlUrl":64},"references\u002Fisolation.md",3645,"https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fageval-config-package\u002Freferences\u002Fisolation.md",[66,72,79,86,93,100],{"skillNo":67,"name":68,"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":69,"categories":70,"keywords":71,"publishedAt":19},"01M7A01JBSS6ZMX56G1MN8WQNY","echo","https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fplugins\u002Fsample-echo\u002Fskills\u002Fecho\u002FSKILL.md",[],[],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":19},"01M7A01DZ4Y8Q2PAJ2RR905BWB","verify-ageval","Verify ageval by driving the public CLI: doctor, a lock that rejects a bad profile, and one minimal-demo task on local plus openai-http against the in-repo loopback gateway. Use when checking lock\u002Frun, seal invariants, or whether a credential skip was reported honestly. Triggers: verify ageval, loopback gateway, AGEVAL_SKIP_REAL_ACP, not run.","https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fverify-ageval\u002FSKILL.md",[],[],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":19},"01M7A019X1VBJF3VZZZW1BTWJP","eval-ops-pitfalls","Pre-run environment checks and suite troubleshooting: dead proxies, invocation timeouts, package mirrors, model routing, environment_image_unresolved, widespread failures, evaluator dependencies, permissions, and CLI\u002Fschema mismatches. Use when preparing an evaluation run or diagnosing acp_timeout, HTTP_PROXY, AGEVAL_AGENT_INVOKE_TIMEOUT, AGEVAL_APT_MIRROR, PEP668, or test collection failures.","https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Feval-ops-pitfalls\u002FSKILL.md",[],[],{"skillNo":87,"name":88,"nameZh":6,"description":89,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":19},"01M7A015RASQWMTH5CD9PJBWJG","ageval-sdk-harness","Write dataset run.py with ageval_sdk (RunContext, Agent\u002FAgentSession, ToolSet, RunTerminal, publish). Use for task run.py, sessions, tool guards, optional evaluator.py LLM-as-judge. Triggers: AgentSession, RunTerminal, ScriptError, ctx.publish_json, write run.py, evaluator.py. SDK never decides PASS or holds host credentials.","https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fageval-sdk-harness\u002FSKILL.md",[],[],{"skillNo":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"publishedAt":19},"01M7A011QP8CFNTWBB1MZEH9XB","ageval-plugin","Author ageval.plugin\u002F1: exclusive and chain slots, registry resolve, environment or executor implementations, image_layers bake. Use for plugin.yaml, attach_stdio hosts, trajectory_collect, README capability and parameter tables. Not an app store. Never invent slots.","https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fageval-plugin\u002FSKILL.md",[],[],{"skillNo":101,"name":102,"nameZh":6,"description":103,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":104,"categories":105,"keywords":106,"publishedAt":19},"01M7A00XTZPNJ16EDPNFQJCA7D","ageval-platform","ageval platform map: authority (docs\u002Fdesign vs ARCHITECTURE vs Issues), Core vs dataset ownership, red lines (no secrets, mechanism-named adapters), evidence grades, which sibling skill to load. Use when entering the repo, asking \"what is ageval\", \"who owns PASS\", \"read order\", or \"evidence grade\". Not a substitute for docs\u002Fdesign.","https:\u002F\u002Fgithub.com\u002FZJU-REAL\u002Fageval\u002Fblob\u002Fmain\u002Fskills\u002Fageval-platform\u002FSKILL.md",[],[]]