Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/melodic-software/claude-code-plugins/video-digestnpx skills add melodic-software/claude-code-plugins --skill video-digestgit clone --depth 1 https://github.com/melodic-software/claude-code-pluginsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/melodic-software/claude-code-plugins/video-digest)<a href="https://agentmods.dev/skills/melodic-software/claude-code-plugins/video-digest"><img src="https://agentmods.dev/badge/skills/melodic-software/claude-code-plugins/video-digest.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00189 | $0.03660 |
| Opus 5 | $0.00095 | $0.01830 |
| Sonnet 5 | $0.00038 | $0.00732 |
| Haiku 4.5 | $0.00019 | $0.00366 |
Grade A, and why
video-digest scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 225 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Pre-computed context
video-extraction deps: !node -e "const fs=require('fs'),path=require('path'),p=process.env.CLAUDE_PLUGIN_DATA;process.stdout.write(p&&fs.existsSync(path.join(p,'node_modules','@melodic','video-digestion'))?'installed':'MISSING - run setup-deps.mjs (see Prerequisites)')"
yt-dlp: !command -v yt-dlp >/dev/null 2>&1 && { yt-dlp --version 2>/dev/null | head -1; :; } || echo "MISSING — install yt-dlp (see Prerequisites)"
ffmpeg: !command -v ffmpeg >/dev/null 2>&1 && { ffmpeg -version 2>/dev/null | head -1; :; } || echo "MISSING — install ffmpeg (watch action only)"
ImageMagick: !command -v magick >/dev/null 2>&1 && { magick -version 2>/dev/null | head -1; :; } || echo "MISSING — install ImageMagick 7 (watch action only)"
Video digest. YouTube and X
Absorb a single public video (transcript + visual frames), harvest reference links, fact-check
claims through deeper external research, and produce a prioritized repo-applicability menu.
Supported sources: YouTube (youtube.com, youtu.be) and X/Twitter status posts (x.com,
twitter.com). An unsupported host fails closed with the supported-source list rather than
guessing.
Durable artifacts live under .work/<watch-epic>/<video-slug>/; the video, bulk frames, working
contact sheets, and shallow clones stay in the OS temp directory. Where this skill says "deeper
research," use whatever external-research capability your project provides, for example the
discovery plugin's /discovery:research / /discovery:research-deep when installed. Treat those
as the reference implementation, not a hard dependency.
Routing
Read a spoke only when its condition holds. These are mutually exclusive by design: a
transcript run loads no watch spoke, and a YouTube run loads no X file. Do not pre-read the set.
| Read | Condition |
|---|---|
reference/sources/youtube.md |
when the URL is a youtube.com / youtu.be video. Accepted shapes, caption ladder, yt-dlp auth & throttle overrides, YouTube failure patterns |
reference/sources/x.md |
when the URL is an x.com / twitter.com status. Canonicalization, 0..N result arity, platform-ASR captions, login-required cases, 429 degradation |
context/watch-pipeline.md |
when running the watch action (or resume into it), the full per-phase procedure |
context/workflow.md |
when you want the watch phase-flow diagram + phase summary table |
context/watch-queue.md |
when running any queue action, or dequeuing via watch / watch <n>. Claim protocol and preflight decision table |
context/output-contract.md |
when about to write or stage slice artifacts, or when a non-default library_dir work root is configured |
context/quality-gates.md |
when grading a phase or a finished slice against binary criteria |
context/synthesis-contract.md |
when harvesting decks, or when applying the frame promotion bar |
context/companion-primary-sources.md |
when the operator supplied companion primary source URL(s) with the queue or watch |
context/gotchas.md |
when a run failed and you need the recovery path |
What ships with it
60 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- context/companion-primary-sources.md 3.6 KB
- context/gotchas.md 1.7 KB
- context/output-contract.md 14 KB
- context/quality-gates.md 11 KB
- context/synthesis-contract.md 2.4 KB
- context/watch-pipeline.md 16 KB
- context/watch-queue.md 8.9 KB
- context/workflow.md 2.3 KB
- evals/evals.json 7.1 KB
- evals/fixtures/driver-video-goldens.json 2.5 KB
- evals/fixtures/x-status-goldens.json 3.9 KB
- extraction/.npmrc 19 B
- extraction/acquisition/acquire-retry-policy.js 1.4 KB runs code
- extraction/acquisition/acquire-retry-policy.test.js 1.5 KB runs code
- extraction/acquisition/acquire-staged.test.js 2.6 KB runs code
- extraction/acquisition/acquire-throttle.js 7.1 KB runs code
- extraction/acquisition/acquire-throttle.test.js 6.7 KB runs code
- extraction/acquisition/acquire-with-retry.js 1.8 KB runs code
- extraction/acquisition/acquire-yt-dlp-auth.js 2.9 KB runs code
- extraction/acquisition/acquire-yt-dlp-auth.test.js 2.7 KB runs code
- extraction/acquisition/acquire.js 11 KB runs code
- extraction/acquisition/acquire.test.js 4.6 KB runs code
- extraction/acquisition/adapter-argv-conformance.test.js 4.2 KB runs code
- extraction/acquisition/build-yt-dlp-args.js 7.8 KB runs code
- extraction/acquisition/build-yt-dlp-args.test.js 7.1 KB runs code
- extraction/acquisition/preflight-metadata.js 11 KB runs code
- extraction/acquisition/preflight-metadata.test.js 7.1 KB runs code
- extraction/acquisition/preflight-null-claim.test.js 1.8 KB runs code
- extraction/acquisition/select-caption.js 5.0 KB runs code
- extraction/acquisition/select-caption.test.js 4.5 KB runs code
- extraction/acquisition/spawn-yt-dlp-with-auth-fallback.js 3.4 KB runs code
- extraction/acquisition/spawn-yt-dlp-with-auth-fallback.test.js 3.1 KB runs code
- extraction/acquisition/video-metadata.js 3.1 KB runs code
- extraction/acquisition/video-metadata.test.js 1.8 KB runs code
- extraction/adapters/adapter-contract.js 22 KB runs code
- extraction/adapters/adapter-contract.test.js 10 KB runs code
- extraction/adapters/dispatch-cli.test.js 1.7 KB runs code
- extraction/adapters/fixture-adapter.test.js 5.6 KB runs code
- extraction/adapters/registry-conformance.test.js 2.7 KB runs code
- extraction/adapters/registry.js 4.5 KB runs code
- extraction/adapters/registry.test.js 3.5 KB runs code
- extraction/adapters/shared-adapter-suite.js 6.9 KB runs code
- extraction/adapters/x.js 38 KB runs code
- extraction/adapters/x.test.js 36 KB runs code
- extraction/adapters/youtube.js 6.3 KB runs code
- extraction/adapters/youtube.test.js 4.1 KB runs code
- extraction/evals/check-research-complete.js 2.6 KB runs code
- extraction/evals/check-research-complete.test.js 1.6 KB runs code
- extraction/evals/check-transcript-goldens.js 3.6 KB runs code
- extraction/evals/check-transcript-goldens.test.js 1.0 KB runs code
- extraction/evals/check-transcript-readability.js 3.2 KB runs code
- extraction/evals/check-watch-outcomes-layout.test.js 3.2 KB runs code
- extraction/evals/check-watch-outcomes.js 25 KB runs code
- extraction/evals/check-watch-outcomes.test.js 3.3 KB runs code
- extraction/evals/check-x-status-goldens.test.js 5.3 KB runs code
- extraction/evals/fixtures/clean-manual-course-snippet.vtt 205 B
- extraction/evals/fixtures/driver-manual-en-snippet.vtt 669 B
- extraction/evals/manual-caption-driver.test.js 2.2 KB runs code
- extraction/harvesting/analyze-harvested-repos.js 4.5 KB runs code
- extraction/harvesting/analyze-harvested-repos.test.js 3.0 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 225 lines · 189 tokens per session scan A 4dd6f9849a94
video-digest is a skill published in the GitHub repository melodic-software/claude-code-plugins (15 stars, last pushed today), licensed MIT. It adds 189 tokens to every session and 3,660 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
solve-constraint-puzzle
For constraint satisfaction: Sudoku, scheduling, N-queens, logic puzzles, SAT-like problems, assignment problems. Uses propagate-then-search pattern.
count-combinations
For probability and counting: permutations, combinations, sample spaces, Monte Carlo simulation, brute-force enumeration, card/dice problems.
career-path-planner
Career goal mapping with skill gap analysis, actionable development plans, and milestone tracking. Use when planning career transitions, identifying skill gaps, setting professional development goals, or evaluating career options.
case-interview-practice
Interactive consulting case interview practice with structured frameworks, feedback mechanisms, and progressive difficulty. Use when preparing for management consulting interviews, case competitions, or business problem-solving exercises.
course-material-creator
Create custom educational materials including syllabi, lesson plans, assessments, reading lists, and learning objectives. Use when building course curricula, training programs, workshops, or educational content.
parallel-orchestrator
Manage parallel Claude Code workstreams using git worktrees. Use when: splitting large tasks across multiple workers, coordinating parallel development, monitoring worker progress, integrating completed work, analyzing work item documents (code reviews, issue lists). Triggers: parallel, orchestrator, worktrees…