fivetaku_fablize
forgecat
A Claude Code harness that makes Opus or any Claude model see multi-step work through with evidence, verification grounding, investigation protocol, per-task routing, and early-stop guard support.
Loading
4 profiles tagged “workflow”
forgecat
A Claude Code harness that makes Opus or any Claude model see multi-step work through with evidence, verification grounding, investigation protocol, per-task routing, and early-stop guard support.
Brainstorm, plan, debug, review, and compound learnings with AI agents
forgecat
AI-DLC adaptive workflow rules profile for project-level agent guidance and stage-by-stage software delivery.
forgecat
Behavioral guidelines to reduce common LLM coding mistakes by making agents surface assumptions, prefer simple implementations, keep edits focused, and define verifiable success criteria.