菜单
Use when a phage-design decision needs literature, database, dataset, gene-essentiality, synteny, viability, bootability, host, threshold, or model evidence.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
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。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/NVIDIA-BioNeMo/bionemo-recipes --skill bionemo-phage-design-research-evidence命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Use when a phage-design workflow must discover or adapt to local GPU, SSH, Slurm, Lepton, manual, or unfamiliar execution infrastructure and produce durable launch, monitoring, resume, or handoff commands.
Use after selecting an Evo 2 phage SFT checkpoint and defining RL objectives to calibrate prompt serialization, temperature, prefix-length distribution, and fixed validation sampling.
Use when an Evo 2 phage SFT project needs a reproducible Microviridae or new phage genome collection from NCBI, paper-linked repositories, supplements, or other public biological databases.
| name | bionemo-phage-design-research-evidence |
| description | Use when a phage-design decision needs literature, database, dataset, gene-essentiality, synteny, viability, bootability, host, threshold, or model evidence. |
| metadata | {"author":"NVIDIA <bionemofeedback@nvidia.com>"} |
Work inside the recipe and result roots selected by the controller. Prefer primary sources and distinguish measured results, author claims, and your inference.
Use the available background sources as needed. The public King phage-generation paper and supplement describe the published experiment, the Black preprint discusses design efficiency and novelty, and the local EMA transcription records a historical draft phage-therapy quality guideline. If an answer is neither common knowledge nor already supported by this skill or its auxiliary notes, consult the relevant full paper, supplement, or guideline rather than guessing. Distinguish historical versions from later publications and current guidance.
For a new target, host, or objective, gather enough evidence to cover complete-genome viability and the productive-infection lifecycle: adsorption and entry, defenses and counter-defenses, replication, morphogenesis and packaging, lysis and progeny, production-host effects, and intended-use safety. Also consider essential/key genes, regulatory architecture, synteny, topology, composition, similarity to viable references, host models, diversity, and final experimental validation. Mark unsupported or non-applicable axes instead of inventing evidence.
Use available life-science skills, public databases, and web sources as useful, following the concise research notes. Do not treat a predictive score, conservation, expression, or transposon depletion as universal essentiality.
For project work, write a concise artifacts/EVIDENCE.md with citations, relevant source versions, methods/results, contradictions, transfer limits, and unresolved decisions, then note the decision in RUNLOG.md. For a read-only question, simply answer with the evidence needed for that question.