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.
git clone --depth 1 https://github.com/dreamrec/TDPilotWrote 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/dreamrec/tdpilot/td-audio-reactive)<a href="https://agentmods.dev/commands/dreamrec/tdpilot/td-audio-reactive"><img src="https://agentmods.dev/badge/commands/dreamrec/tdpilot/td-audio-reactive.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 | $0.00018 | $0.00692 |
| Opus 5 | $0.00009 | $0.00346 |
| Sonnet 5 | $0.00004 | $0.00138 |
| Haiku 4.5 | $0.00002 | $0.00069 |
Grade A, and why
td-audio-reactive 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 7d 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.
What it actually says
Build an audio-reactive TouchDesigner visual whose live audio signal is explicitly bound to a visible TOP parameter. Finish with one low-quality screenshot of the verified stable output. Do not use a legacy macro template.
- Connect and locate. Call
td_get_infoandtd_get_focus. On connection or authentication failure, followtd_sync_diagnoseand stop. Resolve the target root and intended output from the user's request or focused project. - Find a real audio source. If the user supplied a file or microphone,
honor it. Otherwise inspect the target and nearby project scope for existing
audio-file/device CHOPs and sample likely outputs with
td_chop_data. A cooking CHOP with zero signal is not an active source. If no active source or user choice exists, ask the user for an audio file or microphone route and do not build a silent placeholder. - Define the complete intent. Include source, analysis, normalization,
bounded modulation, a visible TOP chain, a stable output, one explicit
CHOP-to-parameter binding, signal/readback checks, nonblack content, and
temporal response. Default the supported binding target to a grounded numeric
levelTOPparameter; do not target MAT, POP, camera, or arbitrary parameters unless the current compiler and grounding explicitly support them. - Ground the concept. Call
td_brain_grounddirectly with the complete intent, source constraint, target root, desired output,include_memory=true, andtrace_level="summary". Read thegrounding_id, authoring contract, availability, parameter semantics, and recalled evidence. - Author and review. Draft the smallest complete CHOP + TOP graph. The
control edge must carry one explicit
chop_reference_expressionbinding with a safe source channel and a registry-backed numeric target parameter. Remove any conflicting static value. Calltd_brain_proposewith the samegrounding_id,draft_schema_version="2", anddetail_level="summary"; repair machine-readable rejections and creatively important stripped parameters. - Execute transactionally. Require complete server-recomputed intent
coverage and zero unresolved semantic edges, then execute the accepted
plan_idwith rollback on failure. - Prove reactivity. Validate the transaction result, final errors, cook health, nonzero source/analysis signal, installed expression or parameter readback, nonblack TOP content, and temporal visual change. Use metadata-only dynamics for time comparison instead of sending two screenshots. Apply at most one assertion-specific repair; otherwise roll back and report the blocker.
- Show it. Take one screenshot of the stable output at
quality=0.25. Explain audio → analysis → normalized control → explicit binding → visual, name two safe live controls, and state the rollback result and any unverified evidence.
Never report an audio-reactive success when the source is silent, the binding is missing, or visual change was not measured.
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.
- 7d ago First seen · 51 lines · 18 tokens per session scan A 179174c7cc2b
td-audio-reactive is a command published in the GitHub repository dreamrec/TDPilot (11 stars, last pushed 1mo ago), licensed MIT. It adds 18 tokens to every session and 692 once invoked, about $0.0001 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 commands, from other repositories
varyshop-carousel
Create a Varyshop multi-slide carousel (LIGHT style, hook → CTA).
claudecode-sounds:soundpack-create
Create a custom soundpack from your own WAV files.
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.