Interpretable Context Methodology is a way to build agent workflows from numbered folders and Markdown files, with each file providing the prompts and context for one stage. It is for sequential tasks where a single AI agent follows a filesystem-defined process and people may review each stage. The catalogue skills support workflows built with this approach.
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 skills add RinDig/Interpretable-Context-Methodology --skill whisper-beat-findergit clone --depth 1 https://github.com/RinDig/Interpretable-Context-MethodologyWrote 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/rindig/interpretable-context-methodology/whisper-beat-finder)<a href="https://agentmods.dev/skills/rindig/interpretable-context-methodology/whisper-beat-finder"><img src="https://agentmods.dev/badge/skills/rindig/interpretable-context-methodology/whisper-beat-finder/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/rindig/interpretable-context-methodology/whisper-beat-finder"><img src="https://agentmods.dev/badge/skills/rindig/interpretable-context-methodology/whisper-beat-finder.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00033 | $0.00772 |
| Opus 5 | $0.00016 | $0.00386 |
| Sonnet 5 | $0.00007 | $0.00154 |
| Haiku 4.5 | $0.00003 | $0.00077 |
Grade A, and why
whisper-beat-finder 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 10d 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 — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
When to Use
Use this skill in Stage 03, immediately after elevenlabs-narration produces an mp3. It outputs two things:
transcript.json-- Whisper's full output with word-level timestampsbeat-timings.md-- absolute timestamps for each beat boundary and key sub-callout
Stage 04 reads beat-timings.md to populate timing.ts and re-time per-scene T constants.
What You Need
- Python with
openai-whisperinstalled - The mp3 produced by Stage 03 (Whisper takes the file path)
- A list of phrases to find (these come from the script, one per beat boundary)
No GPU is required. The default configuration runs medium.en on CPU, which finishes ~1 minute per minute of audio on a modern laptop and is reliable across drivers.
CPU vs GPU
Configured model: {{WHISPER_MODEL}}
We default to medium.en on CPU because large-v3 on GPU has been observed to segfault on some CUDA driver combinations (RTX 5080 + driver 591.86 was one). medium.en is accurate enough for beat extraction and never crashes. See rules/cpu-fallback.md.
If you have a stable GPU setup and need faster turnaround, switch to large-v3 with device="cuda". If anything segfaults, fall back to medium.en on CPU.
How It Works
- Load the model:
whisper.load_model("medium.en", device="cpu") - Transcribe with
word_timestamps=True,language="en",fp16=False - Save the full result to
transcript.json - For each beat in the script, search the word stream for that beat's opening phrase (case-insensitive substring match across N consecutive words). Record the absolute start time.
- Optionally search for sub-callout phrases (mid-beat moments worth animating against) and record their times too.
- Write
beat-timings.mdwith the structured table.
See rules/phrase-matching.md for the matching contract.
Scripts
scripts/transcribe.py-- transcribe an mp3, savetranscript.jsonscripts/find-beats.py-- giventranscript.jsonand a list of beat phrases, emitbeat-timings.md
What ships with it
4 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.
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.
- 10d ago First seen · 63 lines · 33 tokens per session scan A ac3acf903dec
whisper-beat-finder is a skill published in the GitHub repository RinDig/Interpretable-Context-Methodology (1,156 stars, last pushed 1mo ago), licensed MIT. It adds 33 tokens to every session and 772 once invoked, about $0.0002 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 2026-08-30.
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).
diagnostic-stem-delivery
Audio production with diagnostic analysis, timecode parsing from documents, and verified export workflow.
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.