fable-5
01Plugin Claude Code
Fable 5 methodology as a plugin: disciplined decomposition, adversarial verification, plan-change-driven next steps.
Plugin Claude Code
Fable 5 methodology as a plugin: disciplined decomposition, adversarial verification, plan-change-driven next steps.
Plugin Claude Code
Fable 5's working loop for hard tasks: decompose by verification boundaries, verify against the world, choose next by plan-change. Includes scout + refuter agents and a preflight script.
Instructions file CodexOpenCode
AGENTS.md instructions for Learn57130/fable-5: For any hard, multi-step, or ambiguous task, run this loop (full detail in skills/fable-5/SKILL.md).
Hook
Runs when a session starts on startup, resume, clear and compact, running an inline shell check. From Learn57130/fable-5.
Skill Claude CodeCodex
Use when facing a hard, multi-step, or ambiguous task — debugging, architecture, research, large changes — and you need a disciplined loop for breaking it down, checking your work, and choosing the next action instead of improvising.
Agent
Use for fable-5 step 6 — adversarial verification of a conclusion, diff, or analysis before declaring it done. Dispatch with the claim and pointers to the evidence; fresh context is the point, so do not include the reasoning that produced the claim.
Agent
Use for the fable-5 scout pass — read-only comprehension of unfamiliar code or data before decomposing a hard task. Dispatch in parallel (one per subsystem) when the territory is large. Returns a map, not file dumps.