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 soulmachine/skills --skill asus-esc8000-gpu-bios-tuninggit clone --depth 1 https://github.com/soulmachine/skillsWrote 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/soulmachine/skills/asus-esc8000-gpu-bios-tuning)<a href="https://agentmods.dev/skills/soulmachine/skills/asus-esc8000-gpu-bios-tuning"><img src="https://agentmods.dev/badge/skills/soulmachine/skills/asus-esc8000-gpu-bios-tuning/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/soulmachine/skills/asus-esc8000-gpu-bios-tuning"><img src="https://agentmods.dev/badge/skills/soulmachine/skills/asus-esc8000-gpu-bios-tuning.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 | $0.00112 | $0.00935 |
| Opus 5 | $0.00056 | $0.00467 |
| Sonnet 5 | $0.00022 | $0.00187 |
| Haiku 4.5 | $0.00011 | $0.00093 |
Grade A, and why
asus-esc8000-gpu-bios-tuning 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 — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ASUS ESC8000 GPU-Serving BIOS Tuning
Applies a 10-setting low-latency profile + verifies a 4-setting passthrough baseline, over Redfish. Validated on ESC8000-E12P (8× RTX PRO 6000 Blackwell, 2× Xeon 6730P, BIOS 0804). Settings are reversible until you reboot; the tool is dry-run by default.
Quick start
export BMC_PASSWORD='...'
# 1. Dry-run: show what would change on the target (no writes)
python3 scripts/tune_gpu_bios.py --target <BMC_IP>
# 2. Stage only (reversible, no reboot) — settings pending until next boot
python3 scripts/tune_gpu_bios.py --target <BMC_IP> --apply
# 3. Stage + reboot + verify active (the full apply)
python3 scripts/tune_gpu_bios.py --target <BMC_IP> --apply --reboot
Two source modes (pick one)
- Profile (default): targets come from
scripts/gpu_serving_profile.json(token strings validated on BIOS 0804 / registryS0317.8.4.0). - Template (most robust): copy the live values from an already-tuned reference server —
immune to BIOS-version token drift:
Preferpython3 scripts/tune_gpu_bios.py --target <new_BMC_IP> --from-host <tuned_BMC_IP> --apply --reboot--from-hostwhen you have a known-good server (e.g. tuning a second box to match the first).
The profile (what & why)
10 tuned + 4 baseline-verified. Full rationale table in REFERENCE.md. Headline: Resize BAR on, EPP/EPB → Performance, Energy-Efficient-Turbo / C1E-promotion off, Workload → I/O-sensitive, Latency-Optimized on, SNC off (1 NUMA/socket), ASPM-IIO off, AC-loss → Power On. Baseline (must already hold): Above-4G on, SR-IOV on, ACS off (fast GPU P2P — flip to on only at VM adoption). VT-d has no BIOS toggle on this platform (always exposed; gate it OS-side).
Workflow checklist
- Confirm firmware parity first — same BIOS version on source & target, else tokens won't match
(use the
asus-esc8000-firmware-upgradeskill). With--from-host, mismatched versions are flagged. - Dry-run; review the diff with the operator.
--apply --reboot. The reboot isGracefulShutdown → poll Off → On(this build has no GracefulRestart). POST takes ~5–6 min (SNC change forces full memory retraining).- Verify prints
VERIFIED N/N active. The post-reboot host shows0/Nmid-POST — that's normal, it resolves when the host republishes its BIOS map.
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.
- 11d ago First seen · 63 lines · 112 tokens per session scan A 445ec12fe99b
asus-esc8000-gpu-bios-tuning is a skill published in the GitHub repository soulmachine/skills (3 stars, last pushed yesterday), licensed MIT. It adds 112 tokens to every session and 935 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
baoyu-youtube-transcript
A tool for downloading the written captions, subtitles, chapter information, speaker labels, and cover image from a YouTube video using its URL or ID.
baoyu-slide-deck
Generates professional slide deck images from content. Creates outlines with style instructions, then generates individual slide images. Use when user asks to "create slides", "make a presentation", "generate deck", "slide deck", or "PPT".
baoyu-xhs-images
A tool that breaks content into a series of illustrated image cards for Xiaohongshu, a Chinese social platform also called Little Red Book, and similar social posts.
release-skills
A release workflow for preparing and publishing a new version of a software project. A release is a named version that may include updated version files, release notes, tags, or a GitHub Release.
baoyu-comic
A tool for creating original educational comics that explain knowledge or ideas through multiple illustrated panels. It supports different art styles and tones and can create several comics in one batch.
baoyu-cover-image
Generates article cover images with 5 dimensions (type, palette, rendering, text, mood) combining 11 color palettes and 7 rendering styles. Supports cinematic (2.35:1), widescreen (16:9), and square (1:1) aspects. Use when user asks to "generate cover image", "create article cover", or "make cover".