Skip to content
Back to skills

Legion Memory

ASecurity

Auto-triggered skill that reminds agents of the recall-before-grep doctrine. When an agent is about to search the codebase for answers, this skill reminds them to check legion memory first. Use this when the agent is researching a problem, looking for patterns, or trying to understand why something was built a certain way.

  • 3 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added September 23, 2026
ai-agentsgobash

Security analysis

A100/100

Scanned September 23, 2026

npx -y skills add runlegion/legion --skill legion-memory --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Legion Memory?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Legion Memory
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/runlegion-legion-memory/badge)](https://www.skillsdirectory.com/skills/runlegion-legion-memory)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: legion-memory
description: |
  Auto-triggered skill that reminds agents of the recall-before-grep doctrine. When an agent is about to search the codebase for answers, this skill reminds them to check legion memory first. Use this when the agent is researching a problem, looking for patterns, or trying to understand why something was built a certain way.
version: 1.0.0
user-invocable: false
allowed-tools: Bash, Read
---

# Legion Memory: Recall Before Grep

Before searching the codebase, check legion memory. Code tells you WHAT exists. Legion tells you WHY it exists, WHAT went wrong last time, and WHAT the person who solved it wished they had known.

## Order of Operations

1. `legion recall --repo <current-repo> --context "<your problem>"` -- search your own memory
2. `legion consult --context "<your problem>"` -- search ALL agents if recall did not help
3. THEN grep, glob, read the codebase

## When You Find Useful Reflections

Boost them: `legion boost --id <reflection-id>`

This makes them surface higher for the next agent. Every boost improves the system.

## When You Learn Something New

Reflect it: `legion reflect --repo <current-repo> --text "<what you learned>"`

Capture the WHY, not the WHAT. The code already shows what you did. The reflection should capture:
- Why you chose this approach over alternatives
- What you tried that did not work
- What surprised you
- What the next agent should know

## Cross-Agent Knowledge

Use `legion consult` when you hit something outside your domain. Another agent may have already solved it. The knowledge is indexed and ranked -- useful reflections rise, stale ones fade.

## Bullpen Awareness

Check the team board periodically: `legion bullpen --repo <current-repo>`
Respond to signals directed at you.

Post only what the whole team needs. A report, a status, or a work summary is not that --
it goes to the caller who asked for the work, and to no one else. If one agent asked, one
agent is the audience; posting it instead tells every repo on the machine what a single
caller requested, and a board that carries everyone's status stops being read at all. A
finding another agent would act on earns the board. A record of what you did does not.

This applies to a report you received as much as one you wrote: relaying a subagent's
report onto the board is the same broadcast, one layer up.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.