Sources ledger

Skill jperrello/C0BALT_CUT/.claude/skills/sources-ledger

Investigation on using Claude Code to automatically generate profitable YouTube videos.

Install

npx -y skills add jperrello/C0BALT_CUT --skill sources-ledger

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Maintain work/sources.json — the registry of which source videos have been processed into shorts. Scans every work/<id>/ingest.json + the output/<slug>/ folder it produced, recording title, url, duration, the produced shorts (with grades/tiers), current disk footprint, and active|reaped status. Also mirrors a keyed per-source bd memory so future sessions recall what's already been clipped (and can re-ingest from the saved URL). Pure deterministic scan, idempotent, no Claude. The "memory" half of the disk-hygiene pair (reap-source is the "cleanup" half).

SKILL.md

2.5 KB, 581 tokens by cl100k_base, as published. Nobody here has run it

sources-ledger

The registry of what's been clipped. Answers "have I already made shorts from this video?" and "where did a reaped source come from?" without decoding opaque work/<sha1> hashes.

Usage

sources-ledger.sh sync          # rebuild work/sources.json from work/ + output/
sources-ledger.sh record <id>   # upsert one source (pipeline end-of-run / reap)
sources-ledger.sh show          # print the registry

Output: `work/sources.json`

An array, newest-ingested first, one entry per source that has an ingest.json:

{
  "id": "ac3763ad4c",
  "slug": "we-broke-into-mrbeast-s-studio",
  "title": "We Broke Into MrBeast's Studio",
  "url": "https://www.youtube.com/watch?v=y8K6QazBqrY",
  "uploader": null,
  "duration_sec": 5726.174,
  "ingested": "2026-06-01",
  "shorts": [ { "name": "....mp4", "grade": 72, "tier": "GOLD", "path": "output/<slug>/....mp4" } ],
  "shorts_count": 3,
  "status": "active",
  "reaped": null,
  "work_bytes": 2782512345
}

slug is computed identically to start.sh (kebab of the source title, falling back to the work id) so it maps work/<id> → output/<slug>/.
status is reaped once source.mp4 is gone or a .reaped marker exists (written by reap-source); otherwise active.
shorts reads each delivered output/<slug>/*.mp4 and its co-located *.grade.json.

bd memory

record also upserts a keyed memory (--key source-<id>) summarizing the source + how to re-ingest, best-effort (skipped if bd is absent). Keyed, so re-runs update in place rather than duplicating.

Where it runs

Pipeline end-of-run — start.sh calls sources-ledger.sh record <id> after a source's shorts are saved, so the registry is always current.
reap-source calls record <id> after reaping to flip status → reaped, then a final sync.
Standalone — sync backfills the whole registry from existing work/ dirs.