菜单
Build production-style Codex skills from a JSON spec, including SKILL.md, skill.yaml, agents/openai.yaml, references, and runnable scripts. Use when Codex needs to scaffold a new tool-specific skill quickly and consistently with this repository's conventions.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use Nucleotide Transformer v3 for long-context multispecies workflows with Hugging Face Transformers as the primary path, with JAX helper APIs as secondary compatibility, including pre-trained MLM embeddings, post-trained species-conditioned track/annotation inference, dual-mode fine-tuning workflows (prep + full training), length-divisibility checks, and gated-repo troubleshooting. Use when Codex needs to write, fix, explain, or review code or notebooks involving `AutoTokenizer.from_pretrained(..., trust_remote_code=True)`, `AutoModelForMaskedLM`, `AutoModel`, `encode_species`, `num_downsamples`, `keep_target_center_fraction`, NTv3 model names, bigwig/bed outputs, NTv3 fine-tuning notebooks, or NTv3 install/auth issues.
Install, configure, and use Evo 2 for DNA sequence scoring, embeddings, and generation across local GPU inference, Docker, and Nvidia hosted API or NIM deployments. Use when Codex needs to write, fix, explain, or review Evo 2 Python code, shell commands, notebooks, or deployment steps involving checkpoint selection, hardware compatibility, `Evo2(...)`, tokenization, forward passes, embeddings extraction, generation, or Evo 2 installation and troubleshooting.
Build and debug AlphaGenome Python API workflows for variant-effect and track-prediction tasks, including API key setup, package installation, `dna_client` creation, selecting `requested_outputs`, adding `ontology_terms`, plotting results, and troubleshooting environment or response issues. Use when Codex needs to write, fix, explain, or review code and notebooks that use `alphagenome`, `dna_client`, `genome.Interval`, `genome.Variant`, `predict_variant`, `predict_interval`, AlphaGenome plotting helpers, or AlphaGenome API prediction workflows.
Use Calico Borzoi workflows for RNA-seq coverage prediction from DNA, including environment setup, model download, data processing, mini-model training, variant scoring, sequence-interpretation gradients, and SV/STR analysis. Use when Codex needs to write, fix, explain, or review Borzoi commands, scripts, or notebooks from the official `borzoi` tutorials (`make_data`, `train_model`, `score_variants`, `interpret_sequence`, `analyze_sv`) or example QTL notebooks. Prefer this skill for Borzoi repository workflows; use `segment-nt` only for SegmentBorzoi JAX segmentation APIs from Nucleotide Transformer docs.
Use SegmentNT, SegmentEnformer, and SegmentBorzoi for JAX-based genomic segmentation at nucleotide resolution, including model-family selection, `pmap` inference setup, SegmentNT Yarn rescaling, feature-probability extraction, and sequence/token-length troubleshooting. Use when Codex needs to write, fix, explain, or review code or notebooks involving `get_pretrained_segment_nt_model`, `get_pretrained_segment_enformer_model`, `get_pretrained_segment_borzoi_model`, `rescaling_factor`, `hk.transform_with_state`, segmentation logits, or sequence constraints for nucleotide-resolution annotation.
Use DNABERT-2 for DNA sequence embeddings, GUE benchmark evaluation, and supervised fine-tuning on custom classification datasets with the official Hugging Face + `finetune/train.py` workflow. Use when Codex needs to write, fix, explain, or review Python, shell, notebook, or `torchrun` code involving `zhihan1996/DNABERT-2-117M`, `AutoTokenizer`, `AutoModel`, `AutoModelForSequenceClassification`, `trust_remote_code=True`, DNABERT2 dataset formatting (`train.csv`/`dev.csv`/`test.csv`), or DNABERT2 training hyperparameters.
| name | skill-factory |
| description | Build production-style Codex skills from a JSON spec, including SKILL.md, skill.yaml, agents/openai.yaml, references, and runnable scripts. Use when Codex needs to scaffold a new tool-specific skill quickly and consistently with this repository's conventions. |
Use this skill to produce a new tool-specific skill package interactively or from a pre-built JSON spec. Prefer: gather requirements → build spec → generate → validate → enable — in that order.
Determine whether a spec JSON already exists.
.json file, skip to step 3.Build the spec interactively. Ask the user in order:
my-tool.genome-language-model-inference, api-variant-prediction).["environment-setup", "inference", "troubleshooting"].["my-tool", "my_tool_api"].spec.json per references/spec-schema.md before proceeding.Generate the skill package.
python skills-dev/skill-factory/scripts/create_skill_from_spec.py --spec <spec.json> --dry-run--output-root skills-dev to place output under skills-dev/<id>/ instead of skills/<id>/.--dry-run.--no-register if the skill is not yet ready for routing.--overwrite only when intentionally replacing an existing folder.Validate generated artifacts.
<output-root>/<id>/ contains: SKILL.md, skill.yaml, agents/openai.yaml.references/ has four .md files.scripts/run_real_<id>_workflow.py exists and is executable.bash scripts/validate_registry.sh from repo root.bash scripts/validate_skill_metadata.sh from repo root.Enable the skill in the registry.
--no-register was used, run register_skill.py manually (see commands below).registry/skills.yaml has enabled: true for the new skill id.Hand-tune the generated content.
references/inference-patterns.md: replace the TODO block with at least one real executable code snippet.references/constraints.md: replace TODO lines with real limits and coordinate conventions.scripts/run_real_<id>_workflow.py: replace the fake_result stub with real inference or API call logic.SKILL.md (in the generated skill): populate ## Grounded API/CLI Surface with verified import paths and function signatures.python skills-dev/skill-factory/scripts/create_skill_from_spec.py --spec <spec.json>python skills-dev/skill-factory/scripts/create_skill_from_spec.py --spec <spec.json> --dry-runpython skills-dev/skill-factory/scripts/create_skill_from_spec.py --spec <spec.json> --output-root skills-dev ← for dev/experimental skillspython skills-dev/skill-factory/scripts/create_skill_from_spec.py --spec <spec.json> --no-registerpython skills-dev/skill-factory/scripts/create_skill_from_spec.py --spec <spec.json> --overwritepython skills-dev/skill-factory/scripts/create_skill_from_spec.py --list-templatespython skills-dev/skill-factory/scripts/register_skill.py --repo-root . --skill-id <id> --path <output-root>/<id> --family <family> --task <task> --trigger <trigger>A generated skill must include:
SKILL.md — decision flow + grounded API surface, not a READMEskill.yaml — status active, tool contracts, priority rules, input mappingsagents/openai.yaml — display name, short description, default promptreferences/ — four .md docs with real content (no TODO stubs)scripts/run_real_<id>_workflow.py — real inference logic with argparse, output JSON, exit contractreferences/spec-schema.md — field rules and interactive spec-building guideassets/templates/ — all Jinja-style templates used by the generatorscripts/create_skill_from_spec.pyscripts/register_skill.py