link-markdown

name	link-markdown
description	Use when an agent needs to inspect relationships, ix links, dependency edges, or blast radius in Markdown artifacts with quire-cli. Covers quire extract, frontmatter relationship edges, body ix:// link harvest, and practical link-audit workflows.
metadata	{"short-description":"Inspect artifact links"}

Link Markdown

Use this skill to inspect artifact relationships and link blast radius. It is for graph and dependency questions, not prose review.

Extract Edges

Use quire extract when the document has an archetype/module with extraction rules:

quire extract path/to/doc.md --module path/to/module

If the archetype cannot be inferred:

quire extract path/to/doc.md --module path/to/module --archetype FR

Edges appear in .edges and include frontmatter relationships plus body ix:// links harvested by quire-rs.

quire extract path/to/doc.md --module path/to/module | jq '.edges'

Quick Link Audit

For one file:

quire extract path/to/doc.md --module path/to/module \
  | jq -r '.edges[] | "\(.type) \(.target)"'

For many files, loop in the shell and keep the output line-oriented:

for f in spec/**/*.md; do
  quire extract "$f" --module path/to/module \
    | jq -r --arg file "$f" '.edges[] | "\($file)\t\(.type)\t\(.target)"'
done

Blast Radius Questions

For “what depends on X,” extract all edges and filter targets:

jq -r 'select(.[2] == "ix://target") | @tsv'

Use exact targets when possible. If target formats vary, first list unique target strings and normalize only after inspecting examples.

Failure Handling

extract requires a module and an archetype with body_extraction. If it fails because there is no DSL, use quire parse for frontmatter and rg 'ix://' as a fallback, and state that the fallback is not equivalent to quire edge harvest.