Manage task-scoped memory sessions and structured memory layers via a Neo4j-backed CLI.
npx skills add https://github.com/led8/.codex --skill agent-memory-led8
Please help me install this Agent Skill. Skill: agent-memory Source: https://github.com/led8/.codex/tree/main/skills/agent-memory Command: npx skills add https://github.com/led8/.codex --skill agent-memory-led8
This skill provides a structured, shell-driven memory system to manage task-scoped sessions, concise reasoning traces, and review-first durable memory for coding-agent workflows.
Start a local Neo4j instance and create a task-scoped memory session using the memory CLI to begin recording the active task workflow.
High-intent search queries and answers about installing and using this skill.
You can add persistent agent memory using a shell-based workflow that records reasoning traces and durable facts in a Neo4j-backed environment. It coordinates task-scoped sessions with structured short-term, reasoning, and long-term memory layers.
Yes, Neo4j is required to manage task-scoped sessions for CLI agents. The memory system uses a Neo4j-backed environment to store, recall, and manipulate structured knowledge graphs across coding tasks.
Agent memory is structured into three layers: short-term, reasoning, and long-term. You use dedicated CLI commands to capture data, reason about the task, and persist durable knowledge like facts, preferences, and entities.
Yes, you can update and merge entities in the knowledge graph using safe modification paths. The CLI supports replace-fact, replace-preference, alias-entity, and merge-entity workflows to maintain a clean and accurate memory state.
The best way to recall task-specific context is by using the CLI recall and search workflows. These commands retrieve traceable reasoning and durable knowledge from the active task-scoped session stored in the Neo4j memory system.
A shell-based memory system provides traceable reasoning and durable knowledge persistence that stateless agents lack. It ensures reproducible coding sessions by maintaining structured facts and preferences across multiple task workflows.
michaeldowlingStore and retrieve typed memory nodes as a queryable graph in Neo4j.
Lobbi-DocsDeploy the Ahling Command Center intelligence layer with Neo4j, Qdrant, CrewAI, Temporal, and AnythingLLM.
0xUrsanomicsExtracts durable preferences, decisions, and strategic insights from chat transcripts for memory storage.
daffy0208Design, implement, and maintain knowledge graphs for complex entity relationships.
