physicalai-runtime-configuring-inference-pipeline
Configures preprocessors, postprocessors, and runners around InferenceModel via manifest specs and ComponentRegistry. Use when editing physicalai.inference.preprocessors or postprocessors, manifest preprocessor/postprocessor lists, instantiate_component, registered type names, or class_path init_args for inference pipeline components.
How do I install this agent skill?
npx skills add https://github.com/openvinotoolkit/physicalai --skill physicalai-runtime-configuring-inference-pipelineIs this agent skill safe to install?
- Gen Agent Trust Hubpass
This skill provides configuration guidelines for the Physical AI runtime inference pipeline. It outlines how to use manifests to define preprocessors, postprocessors, and runners, and includes security best practices to prevent directory traversal attacks.
- Socketpass
No alerts
- Snykpass
Risk: LOW · No issues
What does this agent skill do?
Configuring the Inference Pipeline
Pipeline order: observation → preprocessors → runner → postprocessors → action output. See docs/how-to/inference/configure-pre-post-processing.md.
Core code:
src/physicalai/inference/component_factory.py—ComponentRegistry,instantiate_component,_MAX_COMPONENT_DEPTH.src/physicalai/inference/model.py— builds processor chains from manifest specs.- Built-ins under
preprocessors/andpostprocessors/; runners underrunners/.
Workflow
-
Read the manifest slice for
preprocessors,postprocessors, andmodel.runner.- Done when: you know whether specs use
type(registry short name) orclass_path.
- Done when: you know whether specs use
-
Prefer
typefor built-ins registered incomponent_factory(e.g. normalize/denormalize patterns in docs). -
Use
class_path+init_argsfor explicit classes:preprocessors: - class_path: physicalai.inference.preprocessors.StatsNormalizer init_args: artifact: stats.safetensors- Done when:
init_argspaths resolve relative to the export directory viaresolve_artifact.
- Done when:
-
Add a new built-in processor:
- Implement subclass of
Preprocessor/Postprocessorin the appropriate package. - Register a short
typename incomponent_factoryif manifest-friendly aliases are needed. - Add unit tests under
tests/unit/inference/preprocessors/orpostprocessors/. - Done when: manifest using
typeorclass_pathinstantiates in a minimalInferenceModeltest.
- Implement subclass of
-
Nested components in
init_argsmust stay within_MAX_COMPONENT_DEPTH; avoid cyclic specs.
Validation loop
uv run pytest tests/unit/inference/preprocessors tests/unit/inference/postprocessors tests/unit/inference/test_manifest.py -q
Required checks
- Processor order matches training/export semantics (normalization before runner, denormalization after).
- Artifact file names in manifests do not traverse paths (
.., absolute paths). - New public processors appear in
docs/reference/inference-api.mdor how-to docs when user-visible. - Runner choice (
SinglePass, chunking runners) stays consistent withpredict_action_chunkvsselect_actiondocs.
References
docs/reference/manifest-schema.mddocs/how-to/inference/use-manifest.md
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/openvinotoolkit/physicalai/physicalai-runtime-configuring-inference-pipeline">View physicalai-runtime-configuring-inference-pipeline on skillZs</a>