AGENT SKILL

deployment-storage

Plan and diagnose Snakemake 9.23.1 software deployment, containers, storage providers, caches, archives, and executor jobscript deployment surfaces.

License 许可
MIT
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
deployment-storage
license
MIT
metadata
{ "disco-role": "operating" }
description
Plan and diagnose Snakemake 9.23.1 software deployment, containers, storage providers, caches, archives, and executor jobscript deployment surfaces.
disable-model-invocation
true

02 资源文件 4

03 SKILL.md 正文

Deployment and Storage

Use this sub-skill when a Snakemake workflow needs reproducible software stacks, external storage, cluster/cloud deployment surfaces, source deployment, caches, or archives.

Route here for

  • Choosing --software-deployment-method / --sdm, --use-conda, --use-apptainer, or --use-envmodules for a workflow run.
  • Writing or diagnosing conda:, global conda:, .pin.txt, .post-deploy.sh, container:, containerized:, and envmodules: directives.
  • Planning storage provider usage with --default-storage-provider, --default-storage-prefix, storage: blocks, storage.<tag>(...), local(...), access flags, credentials, or local-copy behavior.
  • Handling --shared-fs-usage, non-shared executor storage/cache/source deployment, --jobscript, --jobname, or job properties embedded in jobscripts.
  • Creating or reviewing workflow archives with --archive, container specs with --containerize, and between-workflow output caching with SNAKEMAKE_OUTPUT_CACHE / --cache.

Start with

Boundaries

  • General CLI construction, target selection, profiles, and core execution planning belong to the root skill or command/execution sub-skills.
  • Python API and plugin authoring details belong to ../python-api-plugins/SKILL.md.
  • Reports, linting, benchmark interpretation, and test debugging belong to ../debugging-reporting/SKILL.md.
  • Language syntax for ordinary rules, wildcards, inputs, and outputs belongs outside this sub-skill unless it directly affects deployment or storage behavior.

Snakemake 9.23.1 cautions

  • Prefer --software-deployment-method conda apptainer (or --sdm conda apptainer) over older mental models that treat deployment flags as unrelated toggles.
  • --use-conda, --use-apptainer / --use-singularity, and --use-envmodules remain accepted convenience flags for the corresponding deployment methods.
  • Do not suggest --reason; it is not a valid Snakemake 9.23.1 flag, although dry-run output still reports reasons for jobs.
  • External tools and plugins are not bundled by Snakemake itself: conda, Apptainer/Singularity, storage plugins, executor plugins, credentials, and cluster/cloud commands must be present in the target runtime.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351