AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

pacsomatic

Prepares and launches nf-core/pacsomatic matched tumor-normal PacBio HiFi genomics workflows from unaligned BAM inputs. Supports samplesheet generation, pinned Nextflow launch artifacts, local checks, LSF/Slurm/PBS Pro/SGE launcher submission, and startup troubleshooting. Use for pacsomatic run preparation and execution, not general short-read somatic analysis or medical imaging PACS.

Original source
K-Dense-AI
License
MIT
Pinned revision
5f2bc0341354fd5590aef76a533ada1e1e5b4e62

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

This skill also requires local files or dependencies. Remote reading alone may not be enough.

Ask your assistant

Use Agentic Memory skills to read k-dense-ai--pacsomatic at revision 5f2bc0341354fd5590aef76a533ada1e1e5b4e62, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 4

references/agent-playbook.md

Operator playbook

  1. Establish one true matched PacBio HiFi tumor/normal pair from acquisition metadata. Require patient and sample IDs; do not invent clinical identifiers.
  2. Select a reference with exact assembly, contig naming and resource provenance. FASTA alone does not supply VEP, SV annotation, clonality or signature resources.
  3. Prepare artifacts with the helper's --dry-run and --use-current-path. Report each warning and separate file/name checks from unperformed BAM, schema, network, container and scheduler checks.
  4. Review samplesheet.csv: tumor status 1, normal status 0, same patient, distinct samples/files. Confirm that optional .pbi indexes correspond to their BAM. Confirm that MM/ML modification tags survived input processing before expecting methylation results.
  5. Review commit pin, params, profiles and generated shell script. Read local infrastructure configuration before launch. --executor slurm by itself does not set the task executor to Slurm. Default config.yaml is descriptive and is not loaded by Python.
  6. Execute only when requested. A new run uses --run; reusing generated artifacts needs --overwrite. If using a local checkout, record git HEAD and any uncommitted diff separately and omit --pipeline-version.
  7. Return the exact artifact/script paths, run type and checked scope, requested pipeline revision, submission job ID if parsed, and one next QC/triage action. Do not label a queued job complete or helper dry-run as pipeline validation.
  8. On failure, preserve the driver log, .nextflow.log, work task logs, params and samplesheet. Do not delete work/cache before deciding whether to resume. Keep secrets and identifiable sequencing metadata out of shared logs/reports.

All example commands in the main skill use repository-root paths. When the skill is installed elsewhere, resolve scripts/run_pacsomatic.py relative to its skill directory; .github/skills/pacsomatic is not this repository's layout.