[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7ASTRASNA5GY66XZ65XYBY2":3,"skill-same:01M7ASTRASNA5GY66XZ65XYBY2":110},{"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":30,"commitSha":31,"skillDirectory":32,"forksCount":14,"repositoryHtmlUrl":33,"repositoryHomepageUrl":6,"ownerHtmlUrl":34,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":35,"install":36,"resources":49},"01M7ASTRASNA5GY66XZ65XYBY2","pufferlib",null,"Version-aware guidance for PufferLib reinforcement-learning environments, vectorization, policies, PuffeRL training, evaluation, and safe checkpoint review. Use when adapting Gymnasium\u002FPettingZoo environments to published PufferLib 3.0.0 or working with the redesigned native 4.0 source line.","nikhilesh-s\u002Fclaude-toolkit","nikhilesh-s","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F123913414?v=4","01M7ARX6NXB6DFY1CCP91599KE","Nikhilesh Suravarjjala","01M7ARX6PFTQKR6ME1K2S4AZ9K",0,"MIT","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: pufferlib\ndescription: Version-aware guidance for PufferLib reinforcement-learning environments, vectorization, policies, PuffeRL training, evaluation, and safe checkpoint review. Use when adapting Gymnasium\u002FPettingZoo environments to published PufferLib 3.0.0 or working with the redesigned native 4.0 source line.\nlicense: MIT\ncompatibility: Bundled CLIs require Python 3.10+ and use only the standard library. Published pufferlib 3.0.0 supports Python >=3.9 but ships as a native-code source archive; current 4.0 source requires Python >=3.10, Torch >=2.9, and an audited CPU\u002FCUDA toolchain. Network, GPU, native builds, environment plug-ins, assets, checkpoints, and external logging are never required by the bundled CLIs.\nallowed-tools: Read Bash Grep Python\nmetadata:\n  version: \"1.1\"\n  skill-author: \"K-Dense Inc.\"\n  last-reviewed: \"2026-07-23\"\n---\n\n# PufferLib\n\nUse PufferLib with an explicit version profile. Upstream currently has two\nincompatible surfaces:\n\n| Profile | Status on 2026-07-23 | Main use |\n|---|---|---|\n| `pufferlib==3.0.0` | Latest stable PyPI release, published 2025-06-23 | Python\u002FGymnasium\u002FPettingZoo emulation, `pufferlib.vector`, Torch PuffeRL |\n| source `4.0` | Upstream default branch; not the latest stable PyPI artifact | Native C Ocean environments, native CUDA trainer, optional Torch fallback |\n\nDo not combine 3.0 imports with 4.0 config\u002FCLI examples. The 4.0 redesign\nremoved the 3.0 `emulation`, `vector`, and `pytorch` modules from the current\npackage tree.\n\n## Safe defaults\n\n1. Start with bundled synthetic, CPU-only, network-free tools.\n2. Do not import an arbitrary environment by dotted path. Bundled tools accept\n   only allowlisted built-ins and slug identifiers.\n3. Do not install or execute an unreviewed environment package, native\n   extension, ROM, map, checkpoint, or pickle file.\n4. Verify official source, immutable revision, licenses, checksums or\n   attestations, and build hooks. Sandbox native builds and first execution.\n5. Cap steps, environments, agents, workers, threads, buffers, memory, disk,\n   render size, and wall time.\n6. Keep training and evaluation environments\u002Fseeds separate.\n7. Default logging to local\u002Fnone. External logging requires explicit opt-in,\n   disclosure acknowledgment, and separate artifact-upload approval.\n8. Never pass W&B or Neptune credentials via CLI, INI, JSON, tags, run names, or\n   logger configuration. Never print them.\n9. Never dump all environment variables or recursively search for `.env`.\n10. Hash checkpoint bytes before trusted, sandboxed loading; metadata inspection\n    is not proof of safety.\n\n## First local checks\n\nAll bundled CLIs are dependency-free and emit strict JSON:\n\n```bash\npython3 scripts\u002Fenv_template.py --help\npython3 scripts\u002Fenv_contract_validator.py\npython3 scripts\u002Fbenchmark_vectorization.py --backend serial\npython3 scripts\u002Ftrain_template.py\npython3 scripts\u002Fvalidate_plan.py\npython3 scripts\u002Frepro_plan.py\n```\n\nDefaults are synthetic, deterministic, bounded, local, CPU-only, no-network,\nand dry-run where training would otherwise occur.\n\n## Installation and provenance\n\n### Published 3.0.0\n\nPyPI supplies only `pufferlib-3.0.0.tar.gz`:\n\n```text\nsha256: 7df3a3e3f5f894d78d2a1f5374097890aec01473183e748abefe4f3faa10eaa9\nRequires-Python: >=3.9\n```\n\nAfter source\u002Fbuild review, create a pinned uv project:\n\n```bash\nuv venv --python 3.11\nuv add --exact --no-sync \"pufferlib==3.0.0\"\nuv lock\nuv sync --frozen\n```\n\nCommit `pyproject.toml` and `uv.lock`; verify the archive digest and every\nresolved dependency. The source build can compile native code and fetch build\nassets, so resolve\u002Fbuild in a sandbox without credentials or sensitive mounts.\nThe uploaded metadata does not pin Torch or CUDA; do not claim a supported CUDA\nmatrix that PyPI does not declare.\n\n### Current 4.0 source\n\nThe reviewed branch head on 2026-07-23 was:\n\n```text\n25647630e1b15330bb3153a5a0d3ff8d234c3acf\n```\n\nPin the commit, not branch `4.0`:\n\n```bash\nuv add --no-sync \\\n  \"pufferlib @ git+https:\u002F\u002Fgithub.com\u002FPufferAI\u002FPufferLib.git@25647630e1b15330bb3153a5a0d3ff8d234c3acf\"\nuv lock\n```\n\nThe current package declares Python `>=3.10` and Torch `>=2.9`. Upstream\nPufferTank currently uses Ubuntu 24.04, Python 3.12, and an NVIDIA CUDA\n13.0.2\u002FcuDNN development image with the `cu130` Torch index, but does not pin\nthe exact Torch wheel or all system packages. Treat it as a reference, not a\ncomplete lock. Never execute a remote installer directly from a pipe.\n\nRead `references\u002Ftraining.md` before any installation or build.\n\n## Environment workflow\n\n### 1. Validate the contract\n\nGymnasium reset returns `(observation, info)`. Step returns:\n\n```python\n(observation, reward, terminated, truncated, info)\n```\n\nValidate spaces, shapes, dtypes, finite rewards, booleans, reset-before-step,\nreset-after-end, seeding, and cleanup. `terminated` is an MDP terminal;\n`truncated` is an external cutoff such as a time limit. Preserve the distinction\nfor bootstrapping and metrics.\n\n```bash\npython3 scripts\u002Fenv_contract_validator.py \\\n  --steps 64 --episodes 8 --seed 42\n```\n\n### 2. Adapt only after review\n\nPublished 3.0 uses explicit wrappers:\n\n```python\nimport pufferlib.emulation\n\nwrapped = pufferlib.emulation.GymnasiumPufferEnv(reviewed_gymnasium_instance)\n```\n\nFor a reviewed PettingZoo Parallel environment:\n\n```python\nwrapped = pufferlib.emulation.PettingZooPufferEnv(reviewed_parallel_instance)\n```\n\nThere is no supported 3.0 `pufferlib.emulate(...)` shortcut matching the old\nskill. Read `references\u002Fenvironments.md` and `references\u002Fintegration.md`.\n\n### 3. Native environments\n\nPublished 3.0 `PufferEnv` requires\n`single_observation_space`, `single_action_space`, and `num_agents` before\n`super().__init__(buf)`. It uses in-place vector buffers and returns separate\nterminal\u002Ftruncation arrays plus a list of info dictionaries.\n\nCurrent 4.0 uses C bindings. Start from upstream `ocean\u002Fsquared` (single-agent)\nor `ocean\u002Ftarget` (multi-agent), build one environment in local\u002Fsanitized mode,\nand verify every buffer size\u002Ftype\u002Findex before optimization.\n\n## Vectorization workflow\n\nPublished 3.0:\n\n```python\nimport pufferlib.vector\n\nvecenv = pufferlib.vector.make(\n    reviewed_creator,\n    backend=pufferlib.vector.Serial,\n    num_envs=4,\n    seed=42,\n)\n```\n\nMove to `Multiprocessing` only after serial traces pass. Record\n`num_envs`, `num_workers`, `batch_size`, zero-copy mode, start method, agent\ncount, masks, and actual returned shapes. For multi-agent environments, batch\nlength is based on agent slots, not necessarily `num_envs`.\n\nCurrent 4.0 config instead uses:\n\n```ini\n[vec]\ntotal_agents = 4096\nnum_buffers = 2\nnum_threads = 16\n```\n\nRead `references\u002Fvectorization.md`. Benchmark fixed work with warmup and at least\nthree repeats; report simulation and end-to-end training SPS separately. The\nbundled benchmark measures only its synthetic harness.\n\n## Policy workflow\n\nPublished 3.0 policies are Torch modules sized from\n`single_observation_space`\u002F`single_action_space`. Stable recurrent composition\nuses `encode_observations` and `decode_actions`; structured emulation uses\n`pufferlib.pytorch.nativize_dtype` and `nativize_tensor`.\n\nCurrent 4.0 Torch fallback composes:\n\n```python\npufferlib.models.Policy(encoder=encoder, decoder=decoder, network=network)\n```\n\nIt provides MLP, MinGRU, LSTM, and GRU network choices; `--slowly` selects this\nfallback instead of the native backend. Check output\u002Fstate shapes, masks,\nfinite values, gradients, and eager-versus-compiled behavior. See\n`references\u002Fpolicies.md`.\n\n## Training and evaluation\n\nPublished 3.0 trainer import:\n\n```python\nfrom pufferlib import pufferl\n\ntrainer = pufferl.PuffeRL(train_config, vecenv, policy)\n```\n\nCurrent 4.0 CLI:\n\n```bash\npuffer train ENV_NAME\npuffer eval ENV_NAME --load-model-path EXACT_TRUSTED_PATH\npuffer sweep ENV_NAME\n```\n\nGenerate a plan instead of launching by default:\n\n```bash\npython3 scripts\u002Ftrain_template.py \\\n  --profile pypi-3.0.0 \\\n  --environment synthetic \\\n  --device cpu \\\n  --total-timesteps 10000\n```\n\nValidate a custom strict-JSON plan:\n\n```bash\npython3 scripts\u002Fvalidate_plan.py --root . --config plan.json\n```\n\nThe schema rejects secret-bearing keys, unbounded resources, dotted environment\npaths, invalid vector divisibility, mixed-version options, and coupled\ntrain\u002Feval seeds. See `references\u002Ftraining.md`.\n\n## Logging\n\nPufferLib 3.0 exposes W&B and Neptune; current 4.0 CLI exposes W&B. Both are\noptional external services. They may transmit configuration, metrics, source\nmetadata, hardware telemetry, output, and approved artifacts, with privacy,\nretention, access-control, and cost implications.\n\n- W&B credential: named environment variable `WANDB_API_KEY`.\n- Neptune credential: named environment variable `NEPTUNE_API_TOKEN`.\n- Never put values in arguments\u002Fconfig\u002Flogs.\n- Sanitize config keys before logging.\n- Keep source\u002Fmodel upload off unless explicitly approved.\n\nThe planner requires both:\n\n```bash\npython3 scripts\u002Ftrain_template.py \\\n  --logger wandb \\\n  --enable-external-logging \\\n  --acknowledge-external-disclosure\n```\n\nIt reports only the required variable name and never reads its value.\n\n## Checkpoint workflow\n\nPufferLib 3.0 and the 4.0 Torch fallback use Torch serialization; current native\n4.0 writes opaque `.bin` weights. PyTorch warns that untrusted models are\nprograms and that `torch.load` uses unpickling.\n\n```bash\npython3 scripts\u002Finspect_checkpoint.py checkpoint.pt \\\n  --root . \\\n  --expected-sha256 0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef\n```\n\nThe inspector hashes and classifies only. It does not call `torch.load`, import\npickle\u002FTorch, inspect archive members, or extract files. Verify source, license,\narchitecture, environment revision, sidecar metadata, and checksum before any\nsandboxed load. Never use `latest` in a reproducible evaluation.\n\n## Bundled files\n\n### Scripts\n\n- `scripts\u002Fenv_template.py` — deterministic synthetic Gymnasium-style template.\n- `scripts\u002Fenv_contract_validator.py` — bounded contract and seed checks.\n- `scripts\u002Fbenchmark_vectorization.py` — capped serial\u002Fspawn synthetic benchmark.\n- `scripts\u002Ftrain_template.py` — non-executing 3.0\u002F4.0 training-plan generator.\n- `scripts\u002Fvalidate_plan.py` — strict config\u002Fresource\u002Fsecurity validator.\n- `scripts\u002Finspect_checkpoint.py` — metadata\u002Fhash inspection without deserialization.\n- `scripts\u002Frepro_plan.py` — separate-seed evaluation and benchmark plan.\n\n### References\n\n- `references\u002Fenvironments.md` — Gymnasium, stable PufferEnv, emulation, native C.\n- `references\u002Fvectorization.md` — backends, shapes, start methods, benchmarks.\n- `references\u002Fpolicies.md` — stable\u002Fcurrent policy contracts and state safety.\n- `references\u002Ftraining.md` — installs, config, CLI, PuffeRL, eval, logs, checkpoints.\n- `references\u002Fintegration.md` — migration matrix, third-party and credential safety.\n\n## Dated upstream sources\n\n- [PyPI pufferlib 3.0.0](https:\u002F\u002Fpypi.org\u002Fproject\u002Fpufferlib\u002F3.0.0\u002F) —\n  released 2025-06-23; checked 2026-07-23.\n- [PyPI 3.0.0 metadata](https:\u002F\u002Fpypi.org\u002Fpypi\u002Fpufferlib\u002F3.0.0\u002Fjson) —\n  digest\u002Fdependencies; checked 2026-07-23.\n- [PufferLib official docs](https:\u002F\u002Fpuffer.ai\u002Fdocs.html) — current 4.0 docs;\n  checked 2026-07-23.\n- [PufferLib source](https:\u002F\u002Fgithub.com\u002FPufferAI\u002FPufferLib) — default branch and\n  implementation; checked 2026-07-23.\n- [PufferTank 4.0 Dockerfile](https:\u002F\u002Fgithub.com\u002FPufferAI\u002FPufferTank\u002Fblob\u002F4.0\u002Fpuffertank.dockerfile)\n  — CUDA\u002FPython reference; checked 2026-07-23.\n- [PufferLib 2.0 paper](https:\u002F\u002Fopenreview.net\u002Fforum?id=qRyteMTgn0) —\n  Reinforcement Learning Journal, 2025; use only for its stated benchmarks.\n- [PufferLib compatibility paper](https:\u002F\u002Farxiv.org\u002Fabs\u002F2406.12905) —\n  submitted 2024-06-18; describes an earlier API\u002Fperformance profile.\n","# PufferLib\n\nUse PufferLib with an explicit version profile. Upstream currently has two\nincompatible surfaces:\n\n| Profile | Status on 2026-07-23 | Main use |\n|---|---|---|\n| `pufferlib==3.0.0` | Latest stable PyPI release, published 2025-06-23 | Python\u002FGymnasium\u002FPettingZoo emulation, `pufferlib.vector`, Torch PuffeRL |\n| source `4.0` | Upstream default branch; not the latest stable PyPI artifact | Native C Ocean environments, native CUDA trainer, optional Torch fallback |\n\nDo not combine 3.0 imports with 4.0 config\u002FCLI examples. The 4.0 redesign\nremoved the 3.0 `emulation`, `vector`, and `pytorch` modules from the current\npackage tree.\n\n## Safe defaults\n\n1. Start with bundled synthetic, CPU-only, network-free tools.\n2. Do not import an arbitrary environment by dotted path. Bundled tools accept\n   only allowlisted built-ins and slug identifiers.\n3. Do not install or execute an unreviewed environment package, native\n   extension, ROM, map, checkpoint, or pickle file.\n4. Verify official source, immutable revision, licenses, checksums or\n   attestations, and build hooks. Sandbox native builds and first execution.\n5. Cap steps, environments, agents, workers, threads, buffers, memory, disk,\n   render size, and wall time.\n6. Keep training and evaluation environments\u002Fseeds separate.\n7. Default logging to local\u002Fnone. External logging requires explicit opt-in,\n   disclosure acknowledgment, and separate artifact-upload approval.\n8. Never pass W&B or Neptune credentials via CLI, INI, JSON, tags, run names, or\n   logger configuration. Never print them.\n9. Never dump all environment variables or recursively search for `.env`.\n10. Hash checkpoint bytes before trusted, sandboxed loading; metadata inspection\n    is not proof of safety.\n\n## First local checks\n\nAll bundled CLIs are dependency-free and emit strict JSON:\n\n```bash\npython3 scripts\u002Fenv_template.py --help\npython3 scripts\u002Fenv_contract_validator.py\npython3 scripts\u002Fbenchmark_vectorization.py --backend serial\npython3 scripts\u002Ftrain_template.py\npython3 scripts\u002Fvalidate_plan.py\npython3 scripts\u002Frepro_plan.py\n```\n\nDefaults are synthetic, deterministic, bounded, local, CPU-only, no-network,\nand dry-run where training would otherwise occur.\n\n## Installation and provenance\n\n### Published 3.0.0\n\nPyPI supplies only `pufferlib-3.0.0.tar.gz`:\n\n```text\nsha256: 7df3a3e3f5f894d78d2a1f5374097890aec01473183e748abefe4f3faa10eaa9\nRequires-Python: >=3.9\n```\n\nAfter source\u002Fbuild review, create a pinned uv project:\n\n```bash\nuv venv --python 3.11\nuv add --exact --no-sync \"pufferlib==3.0.0\"\nuv lock\nuv sync --frozen\n```\n\nCommit `pyproject.toml` and `uv.lock`; verify the archive digest and every\nresolved dependency. The source build can compile native code and fetch build\nassets, so resolve\u002Fbuild in a sandbox without credentials or sensitive mounts.\nThe uploaded metadata does not pin Torch or CUDA; do not claim a supported CUDA\nmatrix that PyPI does not declare.\n\n### Current 4.0 source\n\nThe reviewed branch head on 2026-07-23 was:\n\n```text\n25647630e1b15330bb3153a5a0d3ff8d234c3acf\n```\n\nPin the commit, not branch `4.0`:\n\n```bash\nuv add --no-sync \\\n  \"pufferlib @ git+https:\u002F\u002Fgithub.com\u002FPufferAI\u002FPufferLib.git@25647630e1b15330bb3153a5a0d3ff8d234c3acf\"\nuv lock\n```\n\nThe current package declares Python `>=3.10` and Torch `>=2.9`. Upstream\nPufferTank currently uses Ubuntu 24.04, Python 3.12, and an NVIDIA CUDA\n13.0.2\u002FcuDNN development image with the `cu130` Torch index, but does not pin\nthe exact Torch wheel or all system packages. Treat it as a reference, not a\ncomplete lock. Never execute a remote installer directly from a pipe.\n\nRead `references\u002Ftraining.md` before any installation or build.\n\n## Environment workflow\n\n### 1. Validate the contract\n\nGymnasium reset returns `(observation, info)`. Step returns:\n\n```python\n(observation, reward, terminated, truncated, info)\n```\n\nValidate spaces, shapes, dtypes, finite rewards, booleans, reset-before-step,\nreset-after-end, seeding, and cleanup. `terminated` is an MDP terminal;\n`truncated` is an external cutoff such as a time limit. Preserve the distinction\nfor bootstrapping and metrics.\n\n```bash\npython3 scripts\u002Fenv_contract_validator.py \\\n  --steps 64 --episodes 8 --seed 42\n```\n\n### 2. Adapt only after review\n\nPublished 3.0 uses explicit wrappers:\n\n```python\nimport pufferlib.emulation\n\nwrapped = pufferlib.emulation.GymnasiumPufferEnv(reviewed_gymnasium_instance)\n```\n\nFor a reviewed PettingZoo Parallel environment:\n\n```python\nwrapped = pufferlib.emulation.PettingZooPufferEnv(reviewed_parallel_instance)\n```\n\nThere is no supported 3.0 `pufferlib.emulate(...)` shortcut matching the old\nskill. Read `references\u002Fenvironments.md` and `references\u002Fintegration.md`.\n\n### 3. Native environments\n\nPublished 3.0 `PufferEnv` requires\n`single_observation_space`, `single_action_space`, and `num_agents` before\n`super().__init__(buf)`. It uses in-place vector buffers and returns separate\nterminal\u002Ftruncation arrays plus a list of info dictionaries.\n\nCurrent 4.0 uses C bindings. Start from upstream `ocean\u002Fsquared` (single-agent)\nor `ocean\u002Ftarget` (multi-agent), build one environment in local\u002Fsanitized mode,\nand verify every buffer size\u002Ftype\u002Findex before optimization.\n\n## Vectorization workflow\n\nPublished 3.0:\n\n```python\nimport pufferlib.vector\n\nvecenv = pufferlib.vector.make(\n    reviewed_creator,\n    backend=pufferlib.vector.Serial,\n    num_envs=4,\n    seed=42,\n)\n```\n\nMove to `Multiprocessing` only after serial traces pass. Record\n`num_envs`, `num_workers`, `batch_size`, zero-copy mode, start method, agent\ncount, masks, and actual returned shapes. For multi-agent environments, batch\nlength is based on agent slots, not necessarily `num_envs`.\n\nCurrent 4.0 config instead uses:\n\n```ini\n[vec]\ntotal_agents = 4096\nnum_buffers = 2\nnum_threads = 16\n```\n\nRead `references\u002Fvectorization.md`. Benchmark fixed work with warmup and at least\nthree repeats; report simulation and end-to-end training SPS separately. The\nbundled benchmark measures only its synthetic harness.\n\n## Policy workflow\n\nPublished 3.0 policies are Torch modules sized from\n`single_observation_space`\u002F`single_action_space`. Stable recurrent composition\nuses `encode_observations` and `decode_actions`; structured emulation uses\n`pufferlib.pytorch.nativize_dtype` and `nativize_tensor`.\n\nCurrent 4.0 Torch fallback composes:\n\n```python\npufferlib.models.Policy(encoder=encoder, decoder=decoder, network=network)\n```\n\nIt provides MLP, MinGRU, LSTM, and GRU network choices; `--slowly` selects this\nfallback instead of the native backend. Check output\u002Fstate shapes, masks,\nfinite values, gradients, and eager-versus-compiled behavior. See\n`references\u002Fpolicies.md`.\n\n## Training and evaluation\n\nPublished 3.0 trainer import:\n\n```python\nfrom pufferlib import pufferl\n\ntrainer = pufferl.PuffeRL(train_config, vecenv, policy)\n```\n\nCurrent 4.0 CLI:\n\n```bash\npuffer train ENV_NAME\npuffer eval ENV_NAME --load-model-path EXACT_TRUSTED_PATH\npuffer sweep ENV_NAME\n```\n\nGenerate a plan instead of launching by default:\n\n```bash\npython3 scripts\u002Ftrain_template.py \\\n  --profile pypi-3.0.0 \\\n  --environment synthetic \\\n  --device cpu \\\n  --total-timesteps 10000\n```\n\nValidate a custom strict-JSON plan:\n\n```bash\npython3 scripts\u002Fvalidate_plan.py --root . --config plan.json\n```\n\nThe schema rejects secret-bearing keys, unbounded resources, dotted environment\npaths, invalid vector divisibility, mixed-version options, and coupled\ntrain\u002Feval seeds. See `references\u002Ftraining.md`.\n\n## Logging\n\nPufferLib 3.0 exposes W&B and Neptune; current 4.0 CLI exposes W&B. Both are\noptional external services. They may transmit configuration, metrics, source\nmetadata, hardware telemetry, output, and approved artifacts, with privacy,\nretention, access-control, and cost implications.\n\n- W&B credential: named environment variable `WANDB_API_KEY`.\n- Neptune credential: named environment variable `NEPTUNE_API_TOKEN`.\n- Never put values in arguments\u002Fconfig\u002Flogs.\n- Sanitize config keys before logging.\n- Keep source\u002Fmodel upload off unless explicitly approved.\n\nThe planner requires both:\n\n```bash\npython3 scripts\u002Ftrain_template.py \\\n  --logger wandb \\\n  --enable-external-logging \\\n  --acknowledge-external-disclosure\n```\n\nIt reports only the required variable name and never reads its value.\n\n## Checkpoint workflow\n\nPufferLib 3.0 and the 4.0 Torch fallback use Torch serialization; current native\n4.0 writes opaque `.bin` weights. PyTorch warns that untrusted models are\nprograms and that `torch.load` uses unpickling.\n\n```bash\npython3 scripts\u002Finspect_checkpoint.py checkpoint.pt \\\n  --root . \\\n  --expected-sha256 0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef\n```\n\nThe inspector hashes and classifies only. It does not call `torch.load`, import\npickle\u002FTorch, inspect archive members, or extract files. Verify source, license,\narchitecture, environment revision, sidecar metadata, and checksum before any\nsandboxed load. Never use `latest` in a reproducible evaluation.\n\n## Bundled files\n\n### Scripts\n\n- `scripts\u002Fenv_template.py` — deterministic synthetic Gymnasium-style template.\n- `scripts\u002Fenv_contract_validator.py` — bounded contract and seed checks.\n- `scripts\u002Fbenchmark_vectorization.py` — capped serial\u002Fspawn synthetic benchmark.\n- `scripts\u002Ftrain_template.py` — non-executing 3.0\u002F4.0 training-plan generator.\n- `scripts\u002Fvalidate_plan.py` — strict config\u002Fresource\u002Fsecurity validator.\n- `scripts\u002Finspect_checkpoint.py` — metadata\u002Fhash inspection without deserialization.\n- `scripts\u002Frepro_plan.py` — separate-seed evaluation and benchmark plan.\n\n### References\n\n- `references\u002Fenvironments.md` — Gymnasium, stable PufferEnv, emulation, native C.\n- `references\u002Fvectorization.md` — backends, shapes, start methods, benchmarks.\n- `references\u002Fpolicies.md` — stable\u002Fcurrent policy contracts and state safety.\n- `references\u002Ftraining.md` — installs, config, CLI, PuffeRL, eval, logs, checkpoints.\n- `references\u002Fintegration.md` — migration matrix, third-party and credential safety.\n\n## Dated upstream sources\n\n- [PyPI pufferlib 3.0.0](https:\u002F\u002Fpypi.org\u002Fproject\u002Fpufferlib\u002F3.0.0\u002F) —\n  released 2025-06-23; checked 2026-07-23.\n- [PyPI 3.0.0 metadata](https:\u002F\u002Fpypi.org\u002Fpypi\u002Fpufferlib\u002F3.0.0\u002Fjson) —\n  digest\u002Fdependencies; checked 2026-07-23.\n- [PufferLib official docs](https:\u002F\u002Fpuffer.ai\u002Fdocs.html) — current 4.0 docs;\n  checked 2026-07-23.\n- [PufferLib source](https:\u002F\u002Fgithub.com\u002FPufferAI\u002FPufferLib) — default branch and\n  implementation; checked 2026-07-23.\n- [PufferTank 4.0 Dockerfile](https:\u002F\u002Fgithub.com\u002FPufferAI\u002FPufferTank\u002Fblob\u002F4.0\u002Fpuffertank.dockerfile)\n  — CUDA\u002FPython reference; checked 2026-07-23.\n- [PufferLib 2.0 paper](https:\u002F\u002Fopenreview.net\u002Fforum?id=qRyteMTgn0) —\n  Reinforcement Learning Journal, 2025; use only for its stated benchmarks.\n- [PufferLib compatibility paper](https:\u002F\u002Farxiv.org\u002Fabs\u002F2406.12905) —\n  submitted 2024-06-18; describes an earlier API\u002Fperformance profile.\n",{"name":5,"license":15,"metadata":24,"description":7,"allowed-tools":28,"compatibility":29},{"version":25,"skill-author":26,"last-reviewed":27},"1.1","K-Dense Inc.","2026-07-23","Read Bash Grep Python","Bundled CLIs require Python 3.10+ and use only the standard library. Published pufferlib 3.0.0 supports Python >=3.9 but ships as a native-code source archive; current 4.0 source requires Python >=3.10, Torch >=2.9, and an audited CPU\u002FCUDA toolchain. Network, GPU, native builds, environment plug-ins, assets, checkpoints, and external logging are never required by the bundled CLIs.","nik-claude-toolkit","4feaa2876384b593d9e30d9190bc56985eeeeb8e","skills\u002Fpufferlib","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s","2026-10-22 14:30:58",{"source":8,"skillName":5,"commands":37,"prompt":47,"zipAvailable":48},[38,41,44],{"tool":39,"command":40},"npx","npx skills add nikhilesh-s\u002Fclaude-toolkit --skill pufferlib",{"tool":42,"command":43},"bunx","bunx skills add nikhilesh-s\u002Fclaude-toolkit --skill pufferlib",{"tool":45,"command":46},"pnpm","pnpm dlx skills add nikhilesh-s\u002Fclaude-toolkit --skill pufferlib","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002FSKILL.md",true,[50,56,60,64,68,72,78,82,86,90,94,98,102,106],{"path":51,"resourceType":52,"resourceTypeDesc":53,"sizeBytes":54,"sourceHtmlUrl":55},"references\u002Fenvironments.md","REFERENCE","参考文档",9340,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Freferences\u002Fenvironments.md",{"path":57,"resourceType":52,"resourceTypeDesc":53,"sizeBytes":58,"sourceHtmlUrl":59},"references\u002Fintegration.md",8876,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Freferences\u002Fintegration.md",{"path":61,"resourceType":52,"resourceTypeDesc":53,"sizeBytes":62,"sourceHtmlUrl":63},"references\u002Fpolicies.md",6962,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Freferences\u002Fpolicies.md",{"path":65,"resourceType":52,"resourceTypeDesc":53,"sizeBytes":66,"sourceHtmlUrl":67},"references\u002Ftraining.md",10424,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Freferences\u002Ftraining.md",{"path":69,"resourceType":52,"resourceTypeDesc":53,"sizeBytes":70,"sourceHtmlUrl":71},"references\u002Fvectorization.md",8402,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Freferences\u002Fvectorization.md",{"path":73,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":76,"sourceHtmlUrl":77},"scripts\u002F__init__.py","SCRIPT","脚本",62,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Fscripts\u002F__init__.py",{"path":79,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":80,"sourceHtmlUrl":81},"scripts\u002F_common.py",6668,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Fscripts\u002F_common.py",{"path":83,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":84,"sourceHtmlUrl":85},"scripts\u002Fbenchmark_vectorization.py",8332,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Fscripts\u002Fbenchmark_vectorization.py",{"path":87,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":88,"sourceHtmlUrl":89},"scripts\u002Fenv_contract_validator.py",6892,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Fscripts\u002Fenv_contract_validator.py",{"path":91,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":92,"sourceHtmlUrl":93},"scripts\u002Fenv_template.py",7251,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Fscripts\u002Fenv_template.py",{"path":95,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":96,"sourceHtmlUrl":97},"scripts\u002Finspect_checkpoint.py",6832,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Fscripts\u002Finspect_checkpoint.py",{"path":99,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":100,"sourceHtmlUrl":101},"scripts\u002Frepro_plan.py",6027,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Fscripts\u002Frepro_plan.py",{"path":103,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":104,"sourceHtmlUrl":105},"scripts\u002Ftrain_template.py",9975,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Fscripts\u002Ftrain_template.py",{"path":107,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":108,"sourceHtmlUrl":109},"scripts\u002Fvalidate_plan.py",18466,"https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fpufferlib\u002Fscripts\u002Fvalidate_plan.py",[111,118,126,133,140,147],{"skillNo":112,"name":113,"nameZh":6,"description":114,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":115,"categories":116,"keywords":117,"publishedAt":19},"01M7AT8339CZMCRMRZP6G1R329","zarr-python","Chunked N-D arrays for cloud storage (Zarr-Python 3). Compressed arrays, parallel I\u002FO, S3\u002FGCS via fsspec, NumPy\u002FDask\u002FXarray compatible, for large-scale scientific computing pipelines.","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fzarr-python\u002FSKILL.md",[],[],{"skillNo":119,"name":120,"nameZh":6,"description":121,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":122,"sourceHtmlUrl":123,"categories":124,"keywords":125,"publishedAt":19},"01M7AT77E7JPH96ETQBNDVVYDM","video-use","Edit any video by conversation. Transcribe, cut, color grade, generate overlay animations, burn subtitles — for talking heads, montages, tutorials, travel, interviews. No presets, no menus. Ask questions, confirm the plan, execute, iterate, persist. Production-correctness rules are hard; everything else is artistic freedom.","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fvideo-use\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fvideo-use\u002FSKILL.md",[],[],{"skillNo":127,"name":128,"nameZh":6,"description":129,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":130,"categories":131,"keywords":132,"publishedAt":19},"01M7AT6YVX000D73Q40SBPZYKP","venue-templates","Prepare journal manuscripts, conference papers, research posters, and grant documents using venue-specific formatting guidance and bundled LaTeX scaffolds. Use when selecting an official template, checking current page or anonymity rules, adapting academic writing to a venue, or inspecting a submission PDF.","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fvenue-templates\u002FSKILL.md",[],[],{"skillNo":134,"name":135,"nameZh":6,"description":136,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":137,"categories":138,"keywords":139,"publishedAt":19},"01M7AT6SGGBENKX1G6V8V9G2MJ","vaex","Use this skill for processing and analyzing large tabular datasets (billions of rows) that exceed available RAM. Vaex excels at out-of-core DataFrame operations, lazy evaluation, fast aggregations, efficient visualization of big data, and machine learning on large datasets. Apply when users need to work with large CSV\u002FHDF5\u002FArrow\u002FParquet files, perform fast statistics on massive datasets, create visualizations of big data, or build ML pipelines that do not fit in memory.","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fvaex\u002FSKILL.md",[],[],{"skillNo":141,"name":142,"nameZh":6,"description":143,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":144,"categories":145,"keywords":146,"publishedAt":19},"01M7AT6HHDJ54MR4HDPFHJYR7G","usfiscaldata","Query the U.S. Treasury Fiscal Data REST API for federal financial data. No API key required. Use for national debt (Debt to the Penny), Daily Treasury Statements, Monthly Treasury Statements, Treasury securities auctions, interest rates, foreign exchange rates, savings bonds, or U.S. government revenue and spending statistics.","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Fusfiscaldata\u002FSKILL.md",[],[],{"skillNo":148,"name":149,"nameZh":6,"description":150,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":151,"categories":152,"keywords":153,"publishedAt":19},"01M7AT662JFE4THBVAZBB3ANJD","uncertainty-and-units","Track physical units and propagate measurement uncertainty in scientific calculations using pint and uncertainties. Use for unit conversion and dimensional checking, GUM uncertainty budgets, Type A and Type B evaluation, coverage factors and expanded uncertainty, Monte Carlo propagation, significant-figure and plus-minus reporting, error propagation through curve fits, CODATA constants, auditing Python code for stripped units or broken uncertainty propagation, and order-of-magnitude plausibility checks using dimensionless groups (Reynolds, Peclet, Damkohler, Knudsen, Biot, Womersley), characteristic scales such as diffusion time or Debye length, and observed magnitude ranges. Trigger on \"is this number physically reasonable\", \"sanity check these units\", \"what regime is this flow in\", or a result that looks off by orders of magnitude.","https:\u002F\u002Fgithub.com\u002Fnikhilesh-s\u002Fclaude-toolkit\u002Fblob\u002Fnik-claude-toolkit\u002Fskills\u002Funcertainty-and-units\u002FSKILL.md",[],[]]