adaptation

Manage SQLite-backed sprint metrics and audit findings for iterative improvement workflows.

Updated May 12, 2026
One-click install
npx skills add https://github.com/FL03/claude-shepherd --skill adaptation-fl03
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: adaptation
Source: https://github.com/FL03/claude-shepherd/tree/main/skills/adaptation
Command: npx skills add https://github.com/FL03/claude-shepherd --skill adaptation-fl03

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This skill prevents recurring failure modes in long-running AI agent sessions by implementing a rigorous harvest-store-inject-cite feedback loop that ensures lessons learned are never forgotten.

Core Features & Use Cases

  • Sprint Metrics Tracking: Automatically captures and grades performance metrics like LOC delta, API usage, and wall-clock time to identify efficiency trends.
  • Insight Taxonomy: Standardizes agent observations into six actionable categories like relocation, duplication, and gap analysis to drive continuous architectural improvement.
  • Excellence Enforcement: Mandates a high-quality bar for all agent work, using automated greps and hooks to block substandard code or duplicated patterns before they land.

Quick Start

Run the adaptation roll command at the end of your sprint to capture performance metrics and harvest lessons for the next cycle.

Frequently Asked Questions about adaptation

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

FAQPage Schema
How do I prevent recurring failure modes in long-running multi-agent workflows?

A persistent self-improvement feedback loop prevents recurring failure modes in multi-agent workflows by harvesting, storing, and injecting prior lessons back into the system prompts of active agents.

What is the best way to track sprint metrics for automated engineering optimization?

Automated sprint metrics tracking captures and grades performance metrics like LOC delta, API usage, and wall-clock time to identify efficiency trends, storing the audit findings in a SQLite-backed registry.

How does automated insight capture enforce code quality in multi-agent systems?

Automated insight capture enforces code quality by standardizing agent observations into actionable categories and integrating with pre-dispatch hooks to block substandard code or duplicated patterns before they land.

Can I use a SQLite-backed registry to manage audit findings for agent flocks?

Yes, managing audit findings for agent flocks can be done using a SQLite-backed registry that systematically identifies and mitigates systemic risks through automated insight capture and prior-lesson injection.

When do I need a systematic insight taxonomy for continuous architectural improvement?

A systematic insight taxonomy is needed for continuous architectural improvement when you want to standardize agent observations into six actionable categories, driving systematic identification and mitigation of systemic risks.

Does integrating pre-dispatch hooks with agent system prompts block duplicated patterns?

Yes, integrating pre-dispatch hooks with agent-level system prompts enforces strict architectural invariants and executes automated greps to block substandard code or duplicated patterns before they land.