forger-retain

Name: forger-retain
Availability: InStock
Author: Ryan

Persists validated claims and failed assumptions into a domain-specific knowledge base.

1|Updated May 19, 2026

Search Tags:#automation #validation #persistence #pipeline #knowledge-base #forger

Ready to run

npx skills add https://github.com/RyanDev1st/FORGER --skill forger-retain

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: forger-retain
Source: https://github.com/RyanDev1st/FORGER/tree/main/skills/forger/phases/retain
Command: npx skills add https://github.com/RyanDev1st/FORGER --skill forger-retain

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill solves the problem of knowledge loss and hallucination in AI agents by ensuring that only verified, tested, and successful outcomes are persisted to a domain-specific knowledge base.

Core Features & Use Cases

Knowledge Persistence: Automatically merges proven claims and failed assumptions into the domain knowledge base after a task is completed.
Self-Evolution: Tracks task success streaks to enable shortcut eligibility, allowing future tasks to bypass redundant search phases.
Use Case: After successfully deploying a new API integration, use this Skill to record the verified error-handling patterns and failed assumptions into the domain knowledge base so future developers avoid the same pitfalls.

Quick Start

Invoke the forger-retain skill to finalize the current workspace and update the domain knowledge base with all validated findings.

Frequently Asked Questions about forger-retain

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

FAQPage Schema

How do I persist validated claims and failed assumptions into a knowledge base?▼

To persist validated knowledge into a knowledge base, this Skill merges proven claims and failed assumptions into a domain-specific ledger. It requires workspace artifacts and telemetry data to perform atomic updates after task completion.

What is the best way to prevent AI agent hallucination and knowledge loss after task completion?▼

Preventing AI agent hallucination requires persisting only verified, tested, and successful outcomes into a domain knowledge base. This ensures future task cycles rely on grounded information rather than untested assumptions.

How do I enable agent self-evolution by tracking task success streaks in my automation pipeline?▼

Enabling agent self-evolution involves tracking task success streaks to unlock shortcut eligibility. This allows future automation pipeline tasks to bypass redundant search phases by referencing retained knowledge.

Does the knowledge persistence process require access to specific workspace artifacts or telemetry data?▼

Yes, knowledge persistence requires access to workspace artifacts, telemetry data, and the domain-specific knowledge ledger. These components are necessary to validate findings and perform atomic updates.

When do I need to finalize my workspace and update the domain knowledge base with validated findings?▼

You need to finalize your workspace and update the domain knowledge base after a task is completed. This operates as the final phase of the FORGER pipeline to ensure only grounded information is retained.