Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add maddexritter-rgb/vibe-editing/plugin install vibe-editingWrote 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/maddexritter-rgb/vibe-editing/audit-captions)<a href="https://agentmods.dev/skills/maddexritter-rgb/vibe-editing/audit-captions"><img src="https://agentmods.dev/badge/skills/maddexritter-rgb/vibe-editing/audit-captions/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/maddexritter-rgb/vibe-editing/audit-captions"><img src="https://agentmods.dev/badge/skills/maddexritter-rgb/vibe-editing/audit-captions.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00080 | $0.01426 |
| Opus 5 | $0.00040 | $0.00713 |
| Sonnet 5 | $0.00016 | $0.00285 |
| Haiku 4.5 | $0.00008 | $0.00143 |
Grade A, and why
audit-captions 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 11d ago.
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 — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.
audit-captions — caption-only quality gate
Fresh-context agent. Receives ONLY the rendered mp4. Does NOT know the transcript, the editing decisions, or the source footage. Tests what a VIEWER would see.
How to run
python3 ${CLAUDE_PLUGIN_ROOT}/skills/audit-captions/scripts/check.py \
--clip 20_DELIVER/v1/clip.mp4 \
--out 10_WORK/audit_captions.json
If the clip has a corresponding .ass subtitle file in 10_WORK/, pass it:
python3 ... --ass 10_WORK/captions.ass
What it checks
1. ACCURACY — do captions match spoken words?
Burnt captions are CHUNK-displayed stylized text: punctuation stripped by design, case folded, numbers digit-formed, verbal fillers and false starts dropped. The comparison is therefore NORMALIZED and WER-gated, not word-exact:
- Layered .ass events (shadow + main) are deduplicated to unique display chunks
- Both sides are normalized (case/punctuation folded, digit ↔ number-word equivalence, safe contraction expansion) and sequence-aligned
- Missing runs of ≤4 consecutive words are stylistic trims (dropped fillers / false starts — by design); runs of ≥5 words are content loss
- FAIL if WER >6% (substitutions + content-loss missing + extra, over spoken words) or any 3+ consecutive-word substitution run (wrong text on screen)
- Token diffs within the threshold are reported as
sample_diffs(ASR noise)
2. SPEAKER ATTRIBUTION — right color on right speaker?
- Speaker count comes from the render metadata (
clip.contract.jsonvia_shared/clip_meta.py); 1 declared speaker → uniform color is CORRECT (pass) - 2+ declared speakers → the dialogue events must use ≥2 distinct primary colors
- FAIL if 2+ speakers all render in one color
3. TIMING SYNC — do captions appear/disappear with speech?
Chunk-display engines legitimately show a sentence-final word in the NEXT chunk (isolated 0.5–0.8s per-word lag at chunk boundaries is by-design), and captions lead by 3–5 frames per SOP. The blocking tests are SYSTEMIC, against each chunk's ALIGNED first word (not nearest-same-word anywhere):
- FAIL if the track median lags speech >350ms or runs >600ms ahead (track-level desync — shifted captions), or >25% of chunks lag their words by >450ms
- FAIL any single chunk >1.2s late or >1.2s early; WARN >450ms late
What ships with it
1 file 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.
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.
- 11d ago First seen · 121 lines · 80 tokens per session scan A 36d16df48786
audit-captions is a skill published in the GitHub repository maddexritter-rgb/vibe-editing (7 stars, last pushed 2mo ago), licensed MIT. It adds 80 tokens to every session and 1,426 once invoked, about $0.0004 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 2026-08-31.
Other skills, from other repositories
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
general-video
Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated…
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
chengfeng-check-updates
An environment manager for a video-editing system. It checks whether its skills and runtime—the software needed to run them—are installed and compatible.
diagnostic-stem-delivery
Audio production with diagnostic analysis, timecode parsing from documents, and verified export workflow.