Forge — Development agent for Claude Code
Claude Code driver agent for Forge Core.
How to install Forge
Installs to ~/.claude/agents/makafuiraymond532-debug-forge-loop-forge.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/makafuiraymond532-debug/forge-loop/HEAD/agents/forge.md -o ~/.claude/agents/makafuiraymond532-debug-forge-loop-forge.md Restart Claude Code, or start a new session, for it to be picked up.
What Forge does
name: forge description: Claude Code driver agent for Forge Core. Tracks coverage/speed/quality, rotates strategies on stagnation, and runs fresh-context evaluation. tools: Read, Write, Edit, Grep, Glob, Bash model: opus
Forge Agent
You are an expert in systematic codebase improvement — coverage maximization, performance optimization, and quality enforcement through structured, KPI-driven iteration.
Core Knowledge
Reference the forge skill for the full protocol: @skills/forge/SK
Alternatives in Development
- Web Quality Skills — Agent Skills for optimizing web quality based on Lighthouse and Core Web Vitals 1.5k ★
- Continuity Specialist — Customer retention and engagement specialist 587 ★
- Librarian Critic — Literature quality critic 250 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Foundry Voss
Use this agent for agent quality calibration, SKILL.md creation and refinement, agent output evaluation, crew
Cell Leader
Intent Cell Architecture leader agent. Receives intent from the driver, creates a team via TeamCreate, spawns
Keepr
Use this agent to evaluate PR review comments and classify them into actionable recommendations. keepr (keeper
Crucible Qg Judge
Stagnation judge for Crucible quality-gate — mechanical cross-round comparison of finding sets. Pinned to Sonn
Verify Debt
Tracks, prioritizes, and manages technical debt. Analyzes code metrics, identifies debt items, detects breakin
Control Quality Adversary
Fresh-context judge of a FLAGGED calibration control — is the control genuinely clean, is its oracle fair, or
Related Skills
Validate Benchmark
Benchmark validation coverage, speed, and evidence quality against baseline metrics
Forge Status
Show Forge Claude Code driver status for this project
Eval Coach
Agent Skill for Evaluation-Driven Development (EDD) - guide AI evaluation strategies with the 50-40-10 rule