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 kunitoki/sonic-skills --skill au-reviewgit clone --depth 1 https://github.com/kunitoki/sonic-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/kunitoki/sonic-skills/au-review)<a href="https://agentmods.dev/skills/kunitoki/sonic-skills/au-review"><img src="https://agentmods.dev/badge/skills/kunitoki/sonic-skills/au-review/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/kunitoki/sonic-skills/au-review"><img src="https://agentmods.dev/badge/skills/kunitoki/sonic-skills/au-review.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 | $0.00119 | $0.01350 |
| Opus 5 | $0.00060 | $0.00675 |
| Sonnet 5 | $0.00024 | $0.00270 |
| Haiku 4.5 | $0.00012 | $0.00135 |
Grade A, and why
au-review 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 — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AudioUnit (AUv2 / AUv3) Review
AUv2 runs in-process inside the host; AUv3 runs in a sandboxed XPC extension. Both share the same realtime-callback constraints, but AUv3 adds ObjC/Swift runtime exposure and app-extension lifecycle rules that introduce a distinct class of violations.
Step 0 — Run universal checks first
Invoke audio-dsp-review and audio-numerics-review before this skill. This skill adds
AU-specific checks on top; it does not replace realtime-safety or numerics reviews.
Step 1 — Identify AU version and structure
Determine which AU variant is under review:
| Variant | Entry point | Render surface |
|---|---|---|
| AUv2 (Component Manager) | AudioComponentRegister / AUMIDIEffectFactory |
AURenderCallback |
| AUv3 (App Extension) | AUAudioUnit subclass, NSExtensionPrincipalClass |
internalRenderBlock |
Locate: the render callback / render block, allocateRenderResourcesAndReturnError:,
deallocateRenderResources, supportedViewConfigurations, and all
kAudioUnitProperty_* / AUParameter access sites.
Step 2 — Scan for AU-specific violations
| Violation | Where to look | Risk |
|---|---|---|
ObjC message send ([obj msg] / .property) in render callback |
AURenderCallback body, internalRenderBlock capture body |
ObjC runtime uses locks and may allocate; not realtime-safe |
Render block calls through self, weak references, Swift closures, or ObjC properties |
AUv3 render block | Runtime calls and weak-reference loads are not realtime-safe |
kAudioUnitProperty_* read/written from render thread |
Any render-path property access | AudioUnitGetProperty / SetProperty are not thread-safe |
| Parameter ramps not applied sample-accurately | AUParameterEvent handling, AUv2 renderProc |
Parameter jumps produce zipper noise on DAW automation |
Tail time not reported via kAudioUnitProperty_TailTime |
Reverb, delay, convolution AUs | Host silences the AU before tail has decayed |
AUv3 extension missing com.apple.security.app-sandbox entitlement |
Xcode target entitlements | App Store / Notarisation rejection; Logic refuses to load |
NSFileManager, NSUserDefaults, or UserDefaults accessed from render block |
AUv3 internalRenderBlock | Main-thread-only / file-I/O APIs; realtime violation and potential deadlock |
AUAudioUnit subclass not calling super in allocateRenderResources |
allocateRenderResourcesAndReturnError: override |
Resources for format negotiation never initialised; silent misrender |
Format negotiation not honouring outputBusses[n].format before allocating |
allocateRenderResources |
Mismatch between agreed and actual format; distortion or crash |
auval failures not addressed |
Overall AU implementation | Plugin rejected or silently disabled by Logic, GarageBand, and MainStage |
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 · 81 lines · 119 tokens per session scan A 798cd06b9d5c
au-review is a skill published in the GitHub repository kunitoki/sonic-skills (19 stars, last pushed 3mo ago), licensed Unlicense. It adds 119 tokens to every session and 1,350 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 2026-08-30.
Other skills, from other repositories
audit-pr
Audit a whole PR against the delivery contract and return MERGE-READY or evidenced blockers with the full URL. Consumes the current review-change REVIEW-PASS receipt instead of re-running review axes; posts a SHA-bound ready comment; never edits or merges. Triggers: "audit-pr", "is this PR ready", "merge gate".
fold-findings
Repair persisted fix-now findings in compatible atomic batches: root-cause fixes, green gate, commit/push, and per-row folded: yes updates. Never reclassify or substitute backlog notes. Triggers: "fold-findings", "fix the review findings", "repair audit blockers".
review-change
Review a change with only applicable internal axes, classify every finding, persist fix-now work, and return one evidence-backed decision. Findings only; --adversarial N uses isolated reviewers; --synthesize fuses supplied reviewer tables. Triggers: "review-change", "review this change", "adversarial review".
review-plan
Independent read-only review of a frozen Engineering plan before execution, in a context that did not cut it: feature or fix snapshot, obligation ledger sweep, phase and validator checks. Returns only PLAN-REVIEW-PASS or PLAN-REVIEW-FAIL with a snapshot-bound receipt (a product-intent gap is PLAN-REVIEW-FAIL with…
review-code
Internal correctness + simplification review pass of the agentic-workflow review pack — composed in-turn by review-change and product-audit; not a menu entry. Checks correctness, error handling, duplication, dead code, and simplification opportunities against the project's own conventions. Findings only; never edits…
review-debt
Internal tech-debt transform pass of the agentic-workflow review pack — composed in-turn by review-change and product-audit; not a menu entry. Transforms the synthesized findings table into explicit debt items, each with a re-trigger condition; it does not rescan the diff. Findings only; never edits code.