wiki-sources-integrity

name	wiki-sources-integrity
description	Protect wiki Sources sections during batch operations. USE THIS SKILL whenever the user is: cleaning up a wiki, running batch edits, removing inbox references, migrating sources, or auditing a vault. Also triggers on "batch cleanup", "remove inbox links", "fix sources", "source audit", or "wiki maintenance".

Wiki Sources Integrity

Maintain clean, permanent Sources sections. Never link to temporary directories.

Prerequisites

• A markdown wiki with a two-tier storage model: inbox/ (temporary) and sources/ (permanent)
• Each person/entity page has a ## Sources section linking to identity files

Workflow

Step 1 — Audit current Sources sections

Find all pages that link to inbox/:

grep -rn "inbox/" wiki/ --include="*.md"

Also identify pages that should have identity file links but do not:

# List all identity files in sources/
find sources/identity/ -type f

# List all pages that currently link to sources/identity/
grep -rl "sources/identity/" wiki/ --include="*.md"

Step 2 — Move inbox content to sources/

For every file referenced from inbox/:

1. Copy the file to the appropriate sources/ subdirectory
2. Update the page's Sources section to point to sources/ instead of inbox/
3. Delete the original from inbox/ only after verifying the copy

Step 3 — Ensure identity files are linked

Every identity file in sources/identity/ (passport scans, ID cards, work permits) must be linked from the corresponding entity page's ## Sources section.

Match files to pages by filename prefix or slug:

# Example: sources/identity/alice-smith-passport.pdf
# Should be linked from wiki/people/alice-smith.md as:
# ![[sources/identity/alice-smith-passport.pdf]]

If a page is missing its identity link, add it:

## Sources

- ![[sources/identity/alice-smith-passport.pdf]]
- ![[sources/identity/alice-smith-id-card.pdf]]

Step 4 — Verify no orphaned Sources sections

After removing inbox links, check that no page's ## Sources section is empty:

grep -A 5 "^## Sources$" wiki/**/*.md | grep -B 1 "^$"

If a Sources section is empty, either:

• Re-add the correct identity file links, or
• Remove the empty ## Sources heading if no sources exist for that page

Step 5 — Run audit and confirm

Check the final state:

# Count of identity files vs. pages linking to them
identity_count=$(find sources/identity/ -type f | wc -l)
linked_count=$(grep -rl "sources/identity/" wiki/ --include="*.md" | wc -l)
echo "Identity files: $identity_count | Pages with identity links: $linked_count"

Both counts should match (or the linked count should be ≥ identity count if some files are shared).

Output Format

After cleanup, present a summary:

Sources Integrity Report
━━━━━━━━━━━━━━━━━━━━━━━━
Inbox links removed: 7
Inbox files migrated to sources/: 7
Identity links added: 3
Identity links already present: 12
Pages with empty Sources sections fixed: 1
Audit: PASSED

Rules

1. Never link to inbox/ from Sources. Inbox is temporary — content gets deleted after distillation. Sources must reference sources/ paths only.
2. Identity files belong in Sources. Passport scans, ID cards, work permits in sources/identity/ must be linked from the corresponding entity page.
3. Contact exports are sources. A contacts.vcf or similar export in sources/identity/ should be linked from every person page that contains data from that file.

Pitfalls

• Batch cleanup can wipe Sources. If a page's Sources section contains ONLY inbox references and you remove them, the entire section may become empty or get deleted. Always check that identity file links survive.
• Shared identity files: One passport scan may belong to multiple pages (e.g., a family member's ID referenced on multiple person pages). Link it from every relevant page.
• Filename matching: Identity files must use a consistent slug prefix ({slug}-passport.pdf) so they can be matched to pages automatically.

References

• references/cleanup-checklist.md — Step-by-step checklist for batch source cleanup