meta/cross-reference-mapper

stars:0
forks:0
watches:0
last updated:N/A

Cross-Reference Mapper

The ORS catalog is a graph: 100+ skills in 18+ categories that should be navigable by clicking between them. This skill produces and maintains the ## Cross-references section in each skill's body.

When to use

  • Adding a new skill — generate its ## Cross-references from the catalog.
  • Adding a new category — propagate the change to all sibling skills.
  • Renaming a skill — find every xref that references it and rewrite.
  • Auditing the catalog — find dangling xrefs (slugs that don't resolve).
  • Answering "what skills should I use for X?" — produce a curated trail.

When NOT to use

  • Reading or using an existing skill (load that skill directly).
  • Auditing for content quality (use meta/integrity-audit).

Hard rules

  1. No fabricated citations. Every cited work must resolve to a verifiable record (DOI, PMID, arXiv ID, ISBN, URL with retrieval date, or stable identifier). If a source cannot be resolved, the skill must say so explicitly rather than presenting the claim as established.
  2. No claim without provenance. Every quantitative or factual claim must point to a file, line, table, figure, dataset, or external source the agent can show. "Trust me" is not provenance.
  3. No silent failure. Every script invocation, API call, or tool use must declare its exit status and what to do on non-zero. A skill that silently swallows errors is a violation.

How xrefs work

Every ## Cross-references section is a flat bullet list:

Example xref section (these slugs are illustrative, not validated):

  ## Cross-references

  - systematic-review — full PRISMA 2020 systematic review with screening.
  - citation-management — Zotero / BibTeX / DOI workflows.
  - meta/integrity-audit — 7-mode AI failure-mode checklist applied before publish.

A slug resolves if either skills/<category>/<slug>/SKILL.md exists or meta/<slug>/SKILL.md exists. The meta/ prefix in the bullet disambiguates.

Generating xrefs for a new skill

When a new SKILL.md is added, generate its ## Cross-references from three sources, in priority order:

1. Same-category siblings

List all other skills in skills/<category>/ with a one-line description taken from each skill's frontmatter description (first sentence).

If the category has only one skill (this one), skip — go to source 2.

2. Tag-overlapping skills

If the new skill has tags ["rna-seq", "alignment"], find all other skills whose tags overlap by ≥1. Score each candidate by:

  • 3 points per shared tag
  • +1 if same category
  • -2 if very different oversight level (one is generative, one is mechanical)
  • +2 if both reference each other in their ## Related Skills or similar

Take the top 5 by score.

3. Workflow-complementary skills

These are skills that handle a later stage of the same pipeline:

  • For a wet-lab protocol skill, list the analysis skills that consume its output.
  • For an analysis skill, list the visualization / writing skills that consume its output.
  • For a writing skill, list the literature-research skills that feed it.

Verifying existing xrefs

Run from the repo root:

python scripts/validate-skills.py skills/   # xrefs are validated by this

A failing xref means the slug was renamed or the skill was moved. To fix:

  1. Identify the renamed/moved skill in the diff.
  2. Find every xref that references the old slug.
  3. Rewrite to use the new slug.

A common cause: a skill is moved from skills/<cat>/old-name/ to skills/<cat>/new-name/ (or to meta/). All xrefs must be updated.

Cross-reference style guide

Good:

- bwa-alignment — BWA-MEM short-read aligner; standard for Illumina DNA-seq.
- samtools-bam-processing — SAM/BAM manipulation; downstream of all aligners.
- star-alignment — spliced aligner for RNA-seq; alternative to BWA for RNA.

Bad:

- bwa-alignment
- samtools-bam-processing
- star-alignment

The one-line description tells the reader whether to click through. A bare slug forces them to load every skill to find the relevant one.

Validation

A skill's xrefs are valid when:

  • Every slug in the section resolves to an existing SKILL.md.
  • Each line has the form - <slug> — <description> (em-dash separator).
  • The skill does NOT xref itself.

Cross-references

  • meta/roadmap-author — use this skill as the final step of authoring.
  • meta/integrity-audit — finds dangling xrefs as part of mode 6 (version drift).
  • meta/release-management — version bumps should trigger an xref re-verification.

Changelog

  • 1.0.0 (2026-06-17) — Initial release. Algorithm distilled from the cross-reference patterns used by the bulk-migration of 116 existing skills.
    Good AI Tools