replication-pack

Packages completed AI experiments from run directories into deterministic ZIP bundles with manifests and SHA-256 checksums.

Updated Apr 25, 2026

Search Tags:#replication #deterministic #reproducibility #ai-analysis #cost-logging #experiment-packaging #environment-snapshot

Ready to run

npx skills add https://github.com/siddvoh/sidds-claude-plugins --skill replication-pack

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: replication-pack
Source: https://github.com/siddvoh/sidds-claude-plugins/tree/main/plugins/research/skills/replication-pack
Command: npx skills add https://github.com/siddvoh/sidds-claude-plugins --skill replication-pack

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Packages a completed AI experiment into a self-contained, reproducible bundle that can be shared and cited.

Core Features & Use Cases

Deterministic packaging: the bundle is self-contained and reproducible without external data.
Comprehensive provenance: includes prompts, model IDs, responses, thinking traces, and cost logs.
Reusable setup: the bundle can be used to reproduce experiments against the listed model configurations.

Quick Start

Trigger the replication pack workflow by providing a run directory or allow the system to select the latest run and run the pack process.

Frequently Asked Questions about replication-pack

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

FAQPage Schema

How do I package an AI experiment into a reproducible bundle?▼

To package an AI experiment into a reproducible bundle, provide a run directory or let the system select the latest run. The workflow collects logs, setup, and environment data to create a deterministic ZIP file.

What is a deterministic replication package for AI analysis?▼

A deterministic replication package is a self-contained bundle that allows an AI experiment to be shared and cited. It includes prompts, model IDs, responses, thinking traces, and cost logs for exact reproducibility.

How does environment snapshotting work for reproducibility?▼

Environment snapshotting works by recording the setup and configuration of your run directory. This data is added to a MANIFEST alongside run metadata and a SHA256 hash to ensure the bundle remains self-contained.

Does the experiment packaging process strip credentials from the bundle?▼

Yes, the experiment packaging process strips credentials from the bundle. It ensures all data is based only on logged records, keeping the shared ZIP file secure while preserving provenance and cost logs.

Can I use this to package runs from a .research directory?▼

Yes, you can package runs from a .research directory, as well as experiments or runs directories. The workflow automatically identifies these locations to collect raw calls and environment data for the bundle.

What limitations exist when sharing an AI experiment bundle?▼

The main limitation is that the bundle is strictly based on logged records. If your raw calls or cost logs are incomplete, the resulting deterministic ZIP will lack that missing provenance data for reproduction.