How do I install the Sources Ledger skill?

Run npx skillmds@latest add jperrello/sources-ledger in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Sources Ledger safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Sources Ledger?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Sources Ledger free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Sources Ledger?

jperrello (@jperrello) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Sources Ledger by jperrello — Agent Skill

Name: Sources Ledger
Author: jperrello

Sources Ledger

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).

jperrello b7ac0c1 3 files · 8.3 KB Updated Sep 17, 2026

File contents

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.

---name: sources-ledgerdescription: 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).---# sources-ledgerThe 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```bashsources-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`:```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.

Frequently asked questions