skills.rest
VMAFx
VMAFxOFFICIAL
Explore all skills by this author →
WebsiteGitHub

ai-run-manifest

Add replay-manifest sidecars to AI-generated artifacts with run provenance.

3|1|Updated May 28, 2026
Search Tags:#auditing#artifacts#manifest#reproducibility#sidecar#ai provenance
One-click install
npx skills add https://github.com/VMAFx/vmafx --skill ai-run-manifest
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ai-run-manifest
Source: https://github.com/VMAFx/vmafx/tree/main/.claude/skills/ai-run-manifest
Command: npx skills add https://github.com/VMAFx/vmafx --skill ai-run-manifest

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you add or audit replay-manifest sidecars so every AI-generated artifact can be traced back to the exact command, inputs, and outputs that created it.

Core Features & Use Cases

  • Manifest Writing: Capture run provenance for dataset builders, feature-table jobs, model exporters, and fetchers.
  • Audit Support: Verify stale artifacts, downloaded caches, and evaluation outputs against their originating scripts and arguments.
  • Use Case: Use it when a script changes schema, defaults, filtering, or output paths and you need a reliable record of what produced the result.

Quick Start

Ask the skill to review the target AI script and specify the manifest sidecar it should write for the artifact.

Frequently Asked Questions about ai-run-manifest

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I track provenance for AI-generated artifacts like datasets and model exports?▼

Replay-manifest sidecars capture run provenance by recording the exact command, inputs, and outputs that generated artifacts like datasets, feature tables, and model exports, ensuring full traceability.

What is a replay-manifest sidecar and when do I need one for my scripts?▼

A replay-manifest sidecar is a provenance file written alongside AI outputs. You need one when scripts change schema, defaults, filtering, or output paths, providing a reliable record of what produced the result.

How do I audit stale artifacts and downloaded caches against their originating scripts?▼

Audit stale artifacts and downloaded caches by verifying them against their originating scripts and arguments using manifest sidecars, which store explicit inputs, outputs, and run provenance for validation.

Does adding manifest sidecars require stable schemas and tests for AI artifact reproducibility?▼

Yes, adding manifest sidecars requires stable manifest schemas, explicit inputs and outputs, run provenance, and tests verifying the sidecar is written, ensuring reliable AI artifact reproducibility.

What's the best way to automatically record run provenance for feature table jobs and fetchers?▼

The best way to record run provenance automatically is to apply manifest sidecars to scripts building feature tables or fetching caches, capturing the exact command, inputs, and arguments that produced the output.

Related Skills

Explore More Software Engineering Skills→
VMAFxVMAFx
3

dev-llm-modelcard

Draft Markdown model cards for ONNX tiny-AI models with provenance validation.

Official
Intermediate
exerias21exerias21

eval-harness

Run pytest tests and pipeline simulations to verify feature behavior.

Community
Intermediate
bcmcpherbcmcpher

datalad-container-run

Automate container provenance capture for DataLad runs.

Community
Advanced
kishorkukrejakishorkukreja
56

track-and-trace

Implement real-time shipment tracking and product traceability systems in Python.

Community
Advanced