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/yhyatt/dmaf/openclaw-skillnpx skills add yhyatt/DMAF --skill openclaw-skillgit clone --depth 1 https://github.com/yhyatt/DMAFWrote 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/yhyatt/dmaf/openclaw-skill)<a href="https://agentmods.dev/skills/yhyatt/dmaf/openclaw-skill"><img src="https://agentmods.dev/badge/skills/yhyatt/dmaf/openclaw-skill.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 | $0.00118 | $0.00826 |
| Opus 5 | $0.00059 | $0.00413 |
| Sonnet 5 | $0.00024 | $0.00165 |
| Haiku 4.5 | $0.00012 | $0.00083 |
Grade A, and why
dmaf 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 5d 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 — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.
DMAF Skill
DMAF watches WhatsApp groups, recognizes faces you care about in photos and videos, and backs them up to Google Photos automatically. Setup requires an agent. After that: zero LLM tokens — the pipeline runs on a system cron + Cloud Run with no AI calls. This skill covers setup (one-time) and day-to-day operations.
Architecture
WhatsApp groups → OpenClaw (auto-download) → GCS bucket → Cloud Run Job → Google Photos
~/.openclaw/media/inbound/ every 30m hourly
Key facts:
- Dedup: Firestore; same file never processed twice (key =
sha256(gs://uri)) - Videos: Sampled at 1–2fps, early exit on first match, full clip uploaded
- Config: YAML stored in Secret Manager (
dmaf-config); update withgcloud secrets versions add - Known people: GCS bucket (
gs://your-project-known-people/); one subdir per person
Phase 1 — First-Time Setup
Read references/setup.md for the full step-by-step guide.
Summary of what you'll create:
- GCP project + required APIs
- Service account
dmaf-runnerwith IAM roles - Two GCS buckets: media staging + known people
- Google Photos OAuth token → Secret Manager
- Config YAML → Secret Manager
- Cloud Build + Cloud Run job
- Cloud Scheduler (hourly trigger)
- Firestore composite indexes
- OpenClaw media sync (system cron)
Ask the user for their GCP project ID before starting. All other values can be derived from that.
Phase 2 — Day-to-Day Operations
Read references/ops.md for full operational commands.
Common tasks:
| Task | Command summary |
|---|---|
| Trigger manual scan | gcloud run jobs execute dmaf-scan --region=us-central1 --async |
| Check scan logs | gcloud logging read 'resource.labels.job_name="dmaf-scan"' --limit=30 --freshness=1h |
| Add a person | Upload photos to gs://your-project-known-people/NAME/, then trigger scan |
| Update config | Edit config YAML, gcloud secrets versions add dmaf-config --data-file=config.yaml |
| Force sync now | bash ~/.openclaw/workspace/scripts/dmaf-sync.sh |
| Check sync log | tail -30 /tmp/dmaf-sync.log |
What ships with it
3 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.
- 5d ago First seen · 75 lines · 0 tokens per session scan A 8879ba6c3dba
dmaf is a skill published in the GitHub repository yhyatt/DMAF (2 stars, last pushed 6mo ago), licensed MIT. It adds 118 tokens to every session and 826 once invoked, about $0.0006 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
performing-kubernetes-etcd-security-assessment
Assess the security posture of Kubernetes etcd clusters by evaluating encryption at rest, TLS configuration, access controls, backup encryption, and network isolation.
Home Security AI Benchmark
LLM & VLM evaluation suite for home security AI applications.
SmartHome Video Anomaly Benchmark
VLM evaluation suite for video anomaly detection in smart home camera footage.
yolo-detection-2026-openvino
OpenVINO — real-time object detection via Docker (NCS2, Intel GPU, CPU).
HomeSafe-Bench
VLM indoor safety hazard detection benchmark inspired by HomeSafeBench (arXiv 2509.23690).
yolo-detection-2026-coral-tpu-win-wsl
Google Coral Edge TPU — real-time object detection natively via Windows WSL.