agentleFS
wheeloswheelos/apollo-lite/.agents/skills/model-artifact-integration/SKILL.md

model-artifact-integration

Skill177 starsChanged 7 days agoApache-2.0View on GitHub

Integrate or replace inference models and engines with verified tensor, preprocessing, runtime, and artifact contracts. Use for ONNX, TensorRT, model assets, precision changes, and target-device deployment.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the owning evaluator/adapter and existing model tests.

    line 10
  2. Distinctive:

    Do not fetch or distribute artifacts without authorization.

    line 12
  3. Distinctive:

    Check invalid shapes, non-finite results, empty inputs, and failure signaling.

    line 20

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Model Artifact Integration

Procedure

  1. Read the owning evaluator/adapter and existing model tests. Identify the model's legitimate source, license, checksum, revision, and target runtime. Do not fetch or distribute artifacts without authorization.
  2. Specify tensor names, shapes, dynamic dimensions, layout, dtype, units, coordinate frames, normalization, preprocessing, and output decoding. Compare training/export semantics with the actual inference implementation.
  3. Record runtime/compiler, device, precision, optimization profiles, plugins, calibration inputs, and engine build parameters. Do not assume a serialized engine is portable across device or runtime versions.
  4. Compare a small reference fixture against independently verified outputs. Check invalid shapes, non-finite results, empty inputs, and failure signaling.
  5. Validate the real target runtime for deployment claims. Use performance-optimization for measured resource claims and simulation-regression for downstream behavioral claims.

Acceptance and failure handling

Separate contract correctness, model quality, target compatibility, and latency evidence. Synthetic plumbing, CPU inference, or engine construction alone does not prove target GPU behavior. Missing device, calibration data, or model rights is a stated boundary, not permission to change precision or substitute assets.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.