vocello-benchmark
Skill388 starsChanged 21 days agoMITView on GitHub
Run or assess a targeted Vocello performance, memory or audio-quality checkpoint using canonical benchmark definitions and compatible evidence.
In this repository8 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- website/CLAUDE.mdCLAUDE.md
- vocello-debugSkill
- vocello-device-diagnosticsSkill
- vocello-ios-validationSkill
- vocello-macos-uiSkill
- vocello-release-readinessSkill
- vocello-ui-autonomySkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 11
Read docs/reference/telemetry-and-benchmarking.md, docs/reference/testing-runbook.md, docs/reference/qc.md and …
- Distinctive: line 25
Never publish benchmark history implicitly.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Benchmark checkpoint
Arguments: $ARGUMENTS
Read docs/reference/telemetry-and-benchmarking.md, docs/reference/testing-runbook.md,
docs/reference/qc.md and docs/reference/tooling-and-evidence.md as relevant. Inspect the benchmark registry and
script help before choosing the smallest applicable scripts/macos_test.sh or scripts/ios_device.sh
lane. Source-bound measurements run on committed checkpoints with complete source identity,
fixed seeds, optimized-build receipts, compatible baselines and required memory qualification.
Join/stop every agent first (SubagentStop) and keep host-wide native serialization. Performance, memory and model
inference/generation lanes retain load/memory preflight. Cache-only rescoring of existing audio/results
uses check_cache_only_host: record load/pressure without blocking, retain ownership/native-lock
checks, and make no latency or memory-qualification claim. If worker
tracking is unavailable explicitly declare lead-only only after joining them. Installed assets only;
large downloads and publication need specific authorization. Do not edit source during a run.
Fast QC gates differ from report-only QC v2 (scripts/qc.py); automated scores do not imply human
listening or supported product-quality claims. Preserve failures; a justified rerun gets a new ID
and explanation. Report actual verdict, source/build identity, baseline compatibility and limits.
Never publish benchmark history implicitly.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

