Auto Memory Refine — Development skill for Claude Code
Run the Critic-Improver refine loop until completion.
How to install Auto Memory Refine
Installs to ~/.claude/skills/hkuds-automemory-auto-memory-refine/SKILL.md
mkdir -p ~/.claude/skills/hkuds-automemory-auto-memory-refine && curl -fsSL https://raw.githubusercontent.com/HKUDS/AutoMemory/HEAD/commands/auto-memory-refine.md -o ~/.claude/skills/hkuds-automemory-auto-memory-refine/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Auto Memory Refine does
name: auto-memory-refine description: Run the Critic-Improver refine loop until completion argument-hint: agent-repo-path alwaysApply: true
auto-memory-refine
**Target agent repository**: $ARGUMENTS
Prerequisites
`/auto-memory-implement` must have completed, leaving `worktraces/refine_state.json` with `step_status: "ready"`.
Goal
Run the full Critic-Improver refine loop: iterate phases until the Critic stops or `max_phases` is reached. This command owns the outer loop and c
Alternatives in Development
- Lazycodex — The one and only agent harness for complex codebases 3.3k ★
- Ralphy — My Ralph Wiggum setup, an autonomous bash script that runs Claude Code, Codex, OpenCode, Cursor agent, Qwen & 2.8k ★
- Claude Code Prompt Improver — Intelligent prompt improver hook for Claude Code 1.3k ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Sdlc Loop
Run the SDLC orchestration loop — auto-dispatch eligible tasks until a gate, blocker, circuit-breaker, or comp
Gauntlet Loop
Fast path for a throwaway proof of concept — set a named outside bar, run a builder against a separate blind c
Goal Loop
Pursue a developer-authored goal contract by iterating build -> verify -> refine until its verifiable criteria
Ralplan
Iterative planning with Planner, Architect, and Critic until consensus
Plan Consensus
Run plan-consensus skill — Claude↔Codex independent-critic review on a plan file until consensus or impasse
Run Daily
Trigger the daily digest now (background via OS scheduler) and monitor until completion
Related Agents
Stp Refiner
Iteratively refine an STP document by running review, fixing findings, and re-reviewing until approved. Autono
Eval Improver
Improver role of the D365FO agent eval loop. Reads the corpus of run records, ranks failure clusters, reproduc
Autoresearch Agent Improver
Improve a target agent.md with a Karpathy-style AutoResearch keep-or-revert loop.