Borrowing it
Nothing to install: this file belongs to chayuto/ws-ESP32-S3-CAM. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/chayuto/ws-ESP32-S3-CAM/main/.claude/commands/audit-wavs.mdgit clone --depth 1 https://github.com/chayuto/ws-ESP32-S3-CAMWrote 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/commands/chayuto/ws-esp32-s3-cam/audit-wavs)<a href="https://agentmods.dev/commands/chayuto/ws-esp32-s3-cam/audit-wavs"><img src="https://agentmods.dev/badge/commands/chayuto/ws-esp32-s3-cam/audit-wavs/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/commands/chayuto/ws-esp32-s3-cam/audit-wavs"><img src="https://agentmods.dev/badge/commands/chayuto/ws-esp32-s3-cam/audit-wavs.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 | $0.00000 | $0.00871 |
| Opus 5 | $0.00000 | $0.00436 |
| Sonnet 5 | $0.00000 | $0.00174 |
| Haiku 4.5 | $0.00000 | $0.00087 |
Grade A, and why
audit-wavs 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 12d 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 — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Audit captured event WAVs
Run the full offline audit pipeline on a directory of cry-detect-01 event WAVs. Produces numeric manifests, per-file spectrogram PNGs, and YAMNet FP32 oracle labels — no human listening required.
Usage: /audit-wavs <dir>
<dir> is a directory laid out as:
<dir>/
├── wavs/ ← 40 s, 16 kHz mono int16 WAVs
├── triggers.jsonl ← optional, ledger from /record/trigger
└── (outputs written here after run)
Typical example: /audit-wavs logs/night-20260418
What the pipeline does
Three stages, orchestrated by projects/cry-detect-01/tools/audit_pipeline.sh:
- Numeric features —
tools/audit_wavs.pymanifest.csv/manifest.jsonl— one row per WAV (32 columns)segments.csv/segments.jsonl— one row per 0.96 s / 0.48 s hop segment (~82 per 40 s file). Time-domain RMS envelope, ZCR, clipping; spectral centroid, rolloff, flatness; 6 energy bands; F0 + HNR.
- Spectrograms —
tools/render_specgrams.pyspecgrams/<stem>.png— mel spectrogram (125 Hz – 7.5 kHz, log axis)- RMS envelope + 0.48 s segment grid. Useful when Claude is doing
the audit: PNGs are readable via the
Readtool.
- RMS envelope + 0.48 s segment grid. Useful when Claude is doing
the audit: PNGs are readable via the
- YAMNet FP32 oracle —
tools/score_yamnet.pyyamnet_files.csv— per-file max/mean of cry-relevant classesyamnet_segments.csv— per 0.96 s frame scores for:baby_cry_infant,crying_sobbing,baby_laughter,speech,child_speech,babbling,screaming,laughterplus the top-3 activated class ids / names.
Segment grids align: segments.csv and yamnet_segments.csv share
(file, seg_idx) identity, so heuristic + oracle columns can be
joined 1:1.
Environment
Uses <repo>/.venv-analysis — Python 3.13 with numpy, scipy,
matplotlib, tensorflow, tensorflow-hub. Create it once with:
python3.13 -m venv .venv-analysis
.venv-analysis/bin/pip install 'setuptools<81' numpy scipy matplotlib tensorflow tensorflow-hub
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.
- 12d ago First seen · 85 lines · 0 tokens per session scan A 4f496c5cd533
audit-wavs is a command published in the GitHub repository chayuto/ws-ESP32-S3-CAM (5 stars, last pushed 2d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 871 tokens. 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 commands, from other repositories
review-firmware
Run embedded firmware code review (memory safety, ISR, RTOS, C/C++ UB).
power-budget
Calculate power budget for the current embedded project.
bom
Generate or analyze Bill of Materials from KiCad project.
debug
Start GDB debug session with OpenOCD or probe-rs.
flash
Build and flash firmware to target device (ESP-IDF or PlatformIO).
ruview-train
Train a RuView model — camera-free WiFlow pose, camera-supervised pose (92.9% PCK@20), RuVector embeddings, domain generalization, local SNN, with optional GPU on GCloud.