Engram Memory Protocol

This is a memory discipline protocol that makes Claude actually remember things across conversations in a codebase. It triggers on decisions, bugfixes, discoveries, and preferences, saving structured context with what, why, where, and what was learned. The skill enforces proactive search before similar work and requires session summaries before closing. What I like here is the specificity: it's not vague "remember stuff" advice, it's a concrete workflow with clear triggers and structured formats. If you're working on a project where Claude keeps forgetting architectural decisions or rediscovering the same bugs, this gives you the scaffolding to build persistent context that survives compaction and session boundaries.

skills.sh installs: 229
GitHub stars: 7.9K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add gentleman-programming/engram --skill engram-memory-protocol --agent claude-code

Installs into .claude/skills of the current project.

Files

When to Use

The agent MUST follow this protocol without waiting for activation. Canonical prose lives in DOCS.md; this contributor skill is manually maintained.

Selective Memory

Persist verified knowledge that is durable and costly to reconstruct: decisions with rationale, non-obvious root causes, real constraints, and resumption-relevant pending work. Group related conclusions deliberately, not every task, response, or event. Exclude secrets, private data, speculation, raw dumps, and routine progress.

Search when materially useful for recall or related work; mere project mentions do not require a search. For recall: mem_context, then mem_search with distinctive keywords, then mem_get_observation for full content. Narrow by project/type after broad recall as needed.

SKILL.md on GitHub

Details

Repository: github.com/gentleman-programming/engram
First seen: Jun 3, 2026