Creates persistent cross-agent memory from scattered coding sessions with episodic, working and procedural layers.
npx skills add https://github.com/Dunc4nJ/agent-skills --skill cm-dunc4nj
Please help me install this Agent Skill. Skill: cm Source: https://github.com/Dunc4nJ/agent-skills/tree/main/skills/cm Command: npx skills add https://github.com/Dunc4nJ/agent-skills --skill cm-dunc4nj
CM transforms scattered coding sessions into a persistent, cross-agent memory so knowledge isn't lost between runs and agents can benefit from shared learnings.
Run cm context "implement memory-safe coding" --json to load relevant rules, history, and suggested queries.
High-intent search queries and answers about installing and using this skill.
To share context across multiple AI coding agents, you need a cross-agent memory system that harmonizes scattered sessions into a persistent knowledge base. This enables discovered patterns and rules to propagate across all agents working on similar tasks.
The best way to prevent repeating mistakes is using a cross-agent memory with procedural memory and safety guardrails. A trauma guard captures past failures, applying confidence decay so outdated or unsafe guidance does not propagate to future agent sessions.
A three-layer cognitive memory architecture works by categorizing agent knowledge into episodic, working, and procedural memory. This structure tracks confidence levels and applies decay over time, ensuring agents retrieve relevant context and rules efficiently during coding tasks.
Yes, you can load relevant history and rules for an AI coding task by running a context command with a natural language query. This action retrieves applicable patterns, historical session data, and suggested follow-up queries to guide the agent.
No, cross-agent memory operates without external dependencies to maintain persistent context. It functions as a standalone architecture that manages episodic, working, and procedural memory layers with built-in confidence tracking and safety guardrails.
AI coding agents lose context between sessions because they lack a persistent cross-agent memory to harmonize scattered sessions. Without a shared memory architecture, discovered strategies and working context are not retained or propagated to future runs.
enowdevIndex agent memory with master and task-specific sub-indexes.
latestaiagentsImplement working, short-term, and long-term memory modules for AI agents.
trungtnmPersist coding agent knowledge across sessions with a Bun/TypeScript CLI.
DicklesworthstoneConsolidate coding sessions into a persistent cross-agent memory store.
