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 google-ai-edge/litert-samples --skill on-device-verificationgit clone --depth 1 https://github.com/google-ai-edge/litert-samplesWrote 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/google-ai-edge/litert-samples/on-device-verification)<a href="https://agentmods.dev/skills/google-ai-edge/litert-samples/on-device-verification"><img src="https://agentmods.dev/badge/skills/google-ai-edge/litert-samples/on-device-verification.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium analysis-evasion · line 1 Suspicious Unicode normalization or mixed-script contentFix: Review the flagged content for security risks. Ensure no credentials, secrets, or sensitive data are exposed.
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 | $0.00099 | $0.02443 |
| Opus 5 | $0.00049 | $0.01222 |
| Sonnet 5 | $0.00020 | $0.00489 |
| Haiku 4.5 | $0.00010 | $0.00244 |
Grade A, and why
on-device-verification 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 8d 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 — 162 lines — stays where its author put it; the contents beside it link to each section on GitHub.
On-device verification
A device result is done when three things hold:
- the model compiles and runs on the accelerator you claim it runs on,
- the output matches the source model numerically and on a task-level gate,
- the record names the device, the runtime version, and the residency line. A number without those is not reproducible and not a result.
Host-side checks (the CompiledModel checker used in gpu-clean-conversion)
exercise the host GPU. The device has its own shader compiler, its own
precision behavior, and its own memory ceiling — every failure mode in the
table below was hit by a model that had already passed on the host.
Loop
1. Dump references from the source model, once. Fixed inputs — one
real sample plus fixed-seed random — saved as .npy next to the recipe
(dump_*_ref.py). These are ground truth for every later step; regenerate
them only when the source model changes.
2. Run the same inputs on the device: CPU first, then GPU. One argument switches the accelerator:
from ai_edge_litert.compiled_model import CompiledModel
from ai_edge_litert.hardware_accelerator import HardwareAccelerator
model = CompiledModel.from_file(
"model.tflite", hardware_accel=HardwareAccelerator.GPU) # or .CPU
The device-CPU run is the control. If it already diverges from the source
dump, the problem is the conversion, not the GPU — go back to
gpu-clean-conversion. A full worked example of the A/B lives in this
repo at samples/litert/speech_recognition/convert/verify_tflite.py.
Ask for the strict accelerator. Compiling with
HardwareAccelerator.CPU | HardwareAccelerator.GPU permits partial
delegation and hides fallback; use the combined mode only to discover
which ops fell back after a strict GPU compile fails.
3. Read the delegate log before reading any numbers.
Replacing N out of M node(s) with delegate ... X partitions
Record N/M and the partition count. N < M or X > 1 means part of the
graph runs on the CPU — decide whether that is acceptable before quoting
any accuracy or latency number.
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.
- 8d ago First seen · 162 lines · 99 tokens per session scan A 4a9209062d80
on-device-verification is a skill published in the GitHub repository google-ai-edge/litert-samples (423 stars, last pushed 4d ago), licensed Apache-2.0. It adds 99 tokens to every session and 2,443 once invoked, about $0.0005 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
guardrails-developer-create-guardrails
Helps developers create a NeMo Guardrails configuration for an LLM application. Use when users want to build, scaffold, configure, test, or iterate on input, output, retrieval, dialog, execution, Colang, or catalog-based guardrails. Trigger keywords - create guardrails, build guardrails, scaffold config, write rails…
cli-eval
Create and run evaluation suites, watch live benchmark progress, view scorecards, compare model performance, and integrate eval runs with CI workflows from the CLI.
model-merging
Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating specialized models by blending domain-specific expertise (math + coding + chat), improving performance beyond single models, or experimenting rapidly with model variants. Covers SLERP, TIES-Merging, DARE, Task…
darwinian-evolver
Evolve prompts/regex/SQL/code with Imbue's evolution loop.
validate
Validate Semantica pipelines, extraction quality, graph schemas, and ontology consistency. Returns structured error/warning checklists. Uses PipelineValidator, PipelineBuilder.validatepipeline(), GraphValidator, and OntologyValidator. Sub-commands: pipeline, step, dependencies, extraction, graph, ontology, performance.
launching-evals
Run, monitor, analyze, and debug LLM evaluations via nemo-evaluator-launcher. Covers running evaluations, checking status and live progress, debugging failed runs, exporting artifacts and logs, and analyzing results. ALWAYS triggers on mentions of running evaluations, checking progress, debugging failed evals…