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-studio-tasksgit 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-studio-tasks)<a href="https://agentmods.dev/skills/fomophil/uad-console-mcp/apollo-studio-tasks"><img src="https://agentmods.dev/badge/skills/fomophil/uad-console-mcp/apollo-studio-tasks/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-studio-tasks"><img src="https://agentmods.dev/badge/skills/fomophil/uad-console-mcp/apollo-studio-tasks.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 | $0.00168 | $0.03225 |
| Opus 5 | $0.00084 | $0.01613 |
| Sonnet 5 | $0.00034 | $0.00645 |
| Haiku 4.5 | $0.00017 | $0.00323 |
Grade A, and why
apollo-studio-tasks 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 — 267 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Apollo studio tasks
Bottom line: read the rig, do one task at a time with the literal sequences below, and check
verified on every write. Channel counts, monitor output indexes, cue bus counts and available
plugins differ per rig. Discover them, never assume them.
Start every session with:
uad_get_status()
uad_list_channels(kind="input")
If safety.mode is readonly, say so up front: nothing can be changed. Exact tool and parameter
names for everything below are in references/tools.md.
Etiquette rails (apply to every task)
- Monitor level: never jump it. The server rails it (default ceiling -6 dB, max +6 dB per call, +10 dB per 30 s) and treats unmute and un-dim as level events. When a call is refused, relay the refusal text verbatim: it names the env var or confirm flag that would unblock it, and the usual fix is to lower the stored level first, then release the mute.
- Never raise the monitor level to check whether something works. Use meters instead.
- 48V: ask what the microphone is before enabling phantom power. It permanently damages ribbon
microphones. Both
phantom_48v=trueandconfirm_condenser_mic=trueare required. - Talkback auto-off:
uad_talkback(on=true)starts a 30 s auto-off timer by default. Calling it again replaces the timer. Never leave talkback open, and never remove the timer to be convenient. - Anything unexpectedly loud, feedback included:
uad_safe_state()first, diagnose second. - Headphone output level cannot be set by this protocol. Build the balance with cue sends and let the performer use the physical knob.
- The server cannot save a Console session. Remind the user to press Cmd+S in Console.
Task: tracking a vocal or an instrument
1. Find the input. It must have has_preamp: true. If more than one candidate fits, ask
rather than guessing.
uad_list_channels(kind="input")
2. Name it and set a clean starting state.
uad_set_channels(items=[
{"target": 1, "name": "LEAD VOX", "fader_db": 0, "mute": false, "solo": false, "hidden": false}
])
What ships with it
5 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 · 267 lines · 168 tokens per session scan A 2c8bef848d16
apollo-studio-tasks is a skill published in the GitHub repository fomoPhil/uad-console-mcp (0 stars, last pushed 1mo ago), licensed MIT. It adds 168 tokens to every session and 3,225 once invoked, about $0.0008 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
tiaportal-mcp
A skill for controlling Siemens TIA Portal, the engineering software used to configure Siemens PLCs and HMIs. It works through an MCP server, a tool connection that lets an agent inspect and change projects.
asset-management
Manage ServiceNow hardware assets, software licenses, and lifecycle states on almhardware/almlicense — license allocation, CMDB-to-asset linking, warranty tracking, inventory aggregation (HAM/SAM).
kernelcad-assemblies
Multi-part assemblies — assembly(), parts, connectors, 7 mate types, fixed and revolute joints, .model()/.solvedModel(). Use for any model with two or more mechanical parts that need joint metadata.
kernelcad-kinematic
Use when verifying whether a moving assembly is buildable — sampled-pose collision sweeps across joint ranges, IK reachability for end-effector targets, mounting-hole fastener consistency, and static-load capacity on cantilever-shaped parts. Loads alongside kernelcad-authoring to gate design-time mechanism feasibility.
kernelcad-parts
Bundled parts catalog — discover, fetch, and mate standard fasteners, bearings, motors, headers, and connectors. Use whenever the model needs an off-the-shelf component instead of hand-modeled placeholder geometry.
use-the-available-kernel
Hard rules for which kernelCAD primitive to reach for when authoring from a reference photo. The from-reference loop fails most often because the author defaults to easy primitives and skips ones that exist. Read before writing any geometry.