Agent memory ledger

Skill Arnie016/codex-prompt-templates/skills/agent-memory-ledger

Create or use lightweight persistent project memory for Codex. Use when the user wants Codex to remember decisions, naming, architecture rules, progress, experiment results, or prior debugging across sessions. Also use with memsearch-style markdown memory.From its SKILL.md

Install

npx -y skills add Arnie016/codex-prompt-templates --skill agent-memory-ledger

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

1.5 KB, 288 tokens by cl100k_base, as published. Nobody here has run it

Agent Memory Ledger

Use markdown as the durable source of truth. Semantic indexes are optional derived caches.

Default Files

Create only the files needed:

.codex-memory/
  decisions.md
  architecture.md
  glossary.md
  experiments.md
  todo.md

Entry Format

## 2026-05-19 - Short Title

- Context:
- Decision:
- Evidence:
- Follow-up:

Recall Workflow

Search .codex-memory/, project docs, and recent task briefs.
If .memsearch/memory/ exists and memsearch is installed, prefer $auto-skill-build-memsearch-codex-memory for progressive recall.
Return a short answer with source file/date.
Update memory only for stable decisions, not every transient observation.

Rules

Do not store secrets.
Do not store full logs unless they are compacted.
Keep each memory file skimmable; archive old details into dated files.
Prefer "why we chose X" over "we talked about X".

Use with $low-token-context for long projects.

Validation

Memory entries cite a source file, command result, or user decision.
Secrets and raw bulky logs are not stored.
A future session can answer "what changed and why?" from the memory files.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.