Compound Learning — Development skill for Claude Code
Compound Learning & Deep Planning.
How to install Compound Learning
Installs to ~/.claude/skills/asklokesh-loki-mode-compound-learning/SKILL.md
mkdir -p ~/.claude/skills/asklokesh-loki-mode-compound-learning && curl -fsSL https://raw.githubusercontent.com/asklokesh/loki-mode/HEAD/skills/compound-learning.md -o ~/.claude/skills/asklokesh-loki-mode-compound-learning/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Compound Learning does
Compound Learning & Deep Planning
**Inspired by:** Compound Engineering Plugin (Every/Kieran Klaassen) -- knowledge compounding philosophy where each unit of work makes subsequent work easier.
Knowledge Compounding (Post-VERIFY)
After VERIFY passes, evaluate whether the task produced a **novel insight** worth preserving. Extract structured solutions that feed back into future planning.
When to Compound
Extract a solution when the task involved:
- Fixing a bug with a non-obviou
Alternatives in Development
- PaperSpine — PaperSpine is a motivation-driven skill for learning from strong academic papers, building a paper’s central a 5k ★
- Claudeception — A Claude Code skill for autonomous skill extraction and continuous learning 2.2k ★
- Cq — An open standard for shared agent learning 1.3k ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Deep Rl
Deep reinforcement learning - DQN/Rainbow/R2D2/Agent57/BBF, PPO/TRPO/GRPO, SAC/TD3/REDQ/DroQ/CrossQ, DreamerV3
DeepTerrainRL
terrain-adaptive locomotion skills using deep reinforcement learning
Twin Sparrow Deep Structure Learning
Enter deep-structure learning mode for books, PDFs, passages, theories, symbols, dreams, religions, scientific
Thinking Critically
Senior staff-level architectural thinking framework. Use when planning systems, evaluating changes, or analyzi
Yggdrasil MCP
Reasoning orchestration MCP server — Tree of Thoughts with deep planning, branching, and revision. Drop-in rep
Obsidian Learn
Review vault learnings, prune stale ones, surface active patterns — the vault's lessons compound or expire
Related Agents
Learning Scribe
Gate-7 capture agent. Converts every deep-review fix or observable incident into a durable docs/learnings/ ent
Polymathic Gates
Reasons through Bill Gates's cognitive architecture — platform thinking, systematic deep-reading, multi-source
Deep Rl Algorithms Curator
Curator for the deep RL algorithms lane of Reinforcement Learning Brain. Use when maintaining source coverage,