redact

Remove specified identifiers from markdown files in the .claude/agent-memory/ directory.

8|1|Updated Jan 24, 2026

Search Tags:#security #memory management #data privacy #agent memory #redact #identifier removal

Ready to run

npx skills add https://github.com/Meme-Theory/meme-engine --skill redact

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: redact
Source: https://github.com/Meme-Theory/meme-engine/tree/main/plugins/research-clab/templates/skills/redact
Command: npx skills add https://github.com/Meme-Theory/meme-engine --skill redact

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill helps maintain data privacy and organization by removing specific references from agent memory, preventing sensitive information from being inadvertently exposed or retained.

Core Features & Use Cases

Targeted Memory Removal: Efficiently delete all mentions of a session, mechanism, or identifier from agent memory files.
Selective Redaction: Option to exclude specific files from the redaction process.
Dry Run Mode: Preview changes before committing to any modifications.
Use Case: After a research session (e.g., session-5) is completed and archived, use /redact session-5 to clean up all associated memory traces, ensuring a tidy and secure agent memory.

Quick Start

Use the redact skill to remove all memory references to "session-12".

Frequently Asked Questions about redact

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I remove specific identifiers from agent memory files?▼

Agent memory redaction works by scanning, filtering, classifying, and editing markdown files within the .claude/agent-memory/ directory to remove specific identifiers, session IDs, or mechanism names based on user-provided flags.

Can I preview memory redaction changes before committing them?▼

Yes, you can preview memory redaction changes by using the dry-run mode, which allows you to review all modifications to agent memory files before any actual edits are committed.

How do I exclude specific files from agent memory redaction?▼

You can exclude specific files from agent memory redaction by using file exclusion flags, ensuring selective redaction targets only the desired markdown files while preserving others in the directory.

When should I redact session IDs from my agent memory footprint?▼

You should redact session IDs from your agent memory footprint when a research session is completed and archived, preventing sensitive information from being inadvertently exposed or retained in memory files.

Does agent memory redaction work with markdown files only?▼

Agent memory redaction operates specifically by scanning and editing markdown files located within the .claude/agent-memory/ directory to ensure data privacy and manage the memory footprint.