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 fomoPhil/uad-console-mcp --skill apollo-onboardinggit clone --depth 1 https://github.com/fomoPhil/uad-console-mcpWrote 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/fomophil/uad-console-mcp/apollo-onboarding)<a href="https://agentmods.dev/skills/fomophil/uad-console-mcp/apollo-onboarding"><img src="https://agentmods.dev/badge/skills/fomophil/uad-console-mcp/apollo-onboarding/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/fomophil/uad-console-mcp/apollo-onboarding"><img src="https://agentmods.dev/badge/skills/fomophil/uad-console-mcp/apollo-onboarding.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 | $0.00183 | $0.03504 |
| Opus 5 | $0.00092 | $0.01752 |
| Sonnet 5 | $0.00037 | $0.00701 |
| Haiku 4.5 | $0.00018 | $0.00350 |
Grade A, and why
apollo-onboarding 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 9d 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 — 255 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Apollo onboarding
Bottom line: discover reality, plan the wiring, settle the clock, then configure. Never assume
per-model I/O. Channel counts, input types and monitor-output indexes differ across Apollo
models. uad_list_channels and uad_raw_get report what is actually there; use them instead of
guessing from the model name.
Rails that apply to every step
- Read before you write.
uad_get_statusanduad_list_channelsfirst, every time. - Never raise the monitor level for the user beyond the rail. The server caps monitor level
at
UAD_MCP_MAX_MONITOR_DB(default -6 dB), allows at most +6 dB per call and +10 dB per 30 s, and treats unmute and un-dim as level events. If a call is refused, relay the refusal text: aSafetyRefusalmessage always names the env var or confirm flag that would unblock it. Do not retry blindly. - Ask before 48V.
uad_set_preamp(phantom_48v=true)also needsconfirm_condenser_mic=true, because phantom power can destroy a ribbon microphone. If the user has not said what the mic is, ask. - Clock and sample-rate changes need
confirm=trueand are refused while audio is flowing (any input above -60 dB). Stop playback and recording first, and say so before changing either. - The server cannot save a Console session. After configuring, remind the user to press
Cmd+S in the Console app.
uad_snapshotis this server's own state file, not a Console session.
Step 0: prove the engine is reachable
uad_get_status()
Note engine.connected, clock.source, clock.locked, clock.sample_rate, every entry in
devices (a cascade shows more than one), and safety.mode. If the mode is readonly, stop and
tell the user that nothing can be changed until UAD_MCP_MODE is set to safe or full.
If the engine is unreachable the error says so plainly: the UA Mixer Engine listens on 127.0.0.1:4710 and starts with the UAD software, staying up after Console is closed.
Step 1: interview the user
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.
- 9d ago First seen · 255 lines · 183 tokens per session scan A 292b099fe72a
apollo-onboarding is a skill published in the GitHub repository fomoPhil/uad-console-mcp (0 stars, last pushed 1mo ago), licensed MIT. It adds 183 tokens to every session and 3,504 once invoked, about $0.0009 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
kernelcad-from-reference
Reproduce a real-world artifact from one or more reference photographs using kernelCAD primitives. Orchestrator skill — names sub-skills and their order. Load this first; it points at the right sub-skill for each stage.
openscad
Create and render OpenSCAD 3D models. Generate preview images from multiple angles, extract customizable parameters, validate syntax, and export STL files for 3D printing platforms like MakerWorld.
seedance-audio
This skill should be used when the user asks for Seedance 2.0 audio, dialogue, lip-sync, music, sound effects, ambience, beat-sync, audio-reference mapping, desync troubleshooting, or sound-driven visual timing.
multimodal-llm
Vision, audio, video generation, and multimodal LLM integration patterns. Use when processing images, transcribing audio, generating speech, generating AI video (Kling v3, Sora 2, Veo 3.1 std/lite/fast, Runway Gen-4.5 via gen4turbo), or building multimodal AI pipelines.
piper-tts-training
Train custom TTS voices for Piper (ONNX format) using fine-tuning or from-scratch approaches. Use when creating new synthetic voices, fine-tuning existing Piper checkpoints, preparing audio datasets for TTS training, or deploying voice models to devices like Raspberry Pi or Home Assistant. Covers dataset preparation…
voice-line
Use when generating TTS voice lines — NPC barks, narrator, dialogue. Covers where voice ids come from, a character-to-voice decision tree, stability/style/similarity guidance, and the multi-line dialogue case via texttodialogue. Trigger on "make a voice line", "narrator", "NPC says", "voice for the merchant", "TTS…