agents-memory
Installation
SKILL.md
Project Memory for Claude Code + Codex
Configure project memory so Claude Code and Codex get stable, scoped instructions across sessions without bloating every prompt. For shared teams, keep the portable project rules in AGENTS.md, mirror that into CLAUDE.md when needed, and keep tool-specific behavior in the tool's native layer.
Treat repo memory as a living exception file. If an agent can reliably infer something by reading the code, config, or README, keep it out of the always-loaded layer. Put only the durable, non-inferable guidance here.
Quick Reference
| Layer | Typical Location | Purpose |
|---|---|---|
| Shared project memory | ./AGENTS.md |
Portable repo instructions for Codex and other AGENTS.md-aware tools |
| Claude project memory | ./CLAUDE.md |
Claude Code project instructions; often a symlink or mirror of AGENTS.md |
| Claude scoped rules | ./.claude/rules/*.md |
Modular Claude-only rules, optionally path-scoped |
| Codex personal memory | ~/.codex/AGENTS.md |
Personal defaults across repositories |
| Codex repo-local override | ./AGENTS.override.md |
Local developer override when you need a non-shared layer |
| Codex auto-memory | ~/.codex/memories/ (opt-in via [features] memories = true) |
Machine-local accumulated recall; off by default in EEA/UK/CH; keep must-always rules in AGENTS.md, not here |
| Claude auto memory | ~/.claude/projects/<project>/memory/ — MEMORY.md index + topic files |
Machine-local, accumulated notes; first 200 lines / 25 KB of MEMORY.md load each session; topic files load on demand |
| Instruction budget | ~100–150 usable lines across all loaded CLAUDE.md tiers | Community-derived heuristic (not an official Anthropic figure): compliance drops past ~150–200 discrete instructions, of which the system prompt already spends ~50; budget is a shared pool across all tiers |
| Layered memory model | references/memory-patterns.md |
Keep the hot memory small; push history and reusable procedures into the right layers |