memory-management

Manages three-tier SEO project memory with hot cache, wiki compilation, and GDPR purge workflows.

Updated May 11, 2026
One-click install
npx skills add https://github.com/cloudofgeorge/AI-hands-Engineer --skill memory-management-cloudofgeorge
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: memory-management
Source: https://github.com/cloudofgeorge/AI-hands-Engineer/tree/main/skills/seo/memory-management
Command: npx skills add https://github.com/cloudofgeorge/AI-hands-Engineer --skill memory-management-cloudofgeorge

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? SEO and GEO projects accumulate findings, decisions, and keyword data across many sessions, and that context is lost between conversations. This Skill maintains a persistent three-tier memory system (HOT/WARM/COLD) so project context survives across sessions without manual note-taking. ## Core Features & Use Cases - Three-Tier Memory Lifecycle: Automatically loads an 80-line hot cache each session, promotes frequently referenced items, demotes stale data, and archives cold storage. - Wiki Compilation Layer: Synthesizes multiple WARM files into compiled wiki pages with source hashes, contradiction reconciliation, and safe retirement of covered files. - GDPR/CCPA Privacy Compliance: Executes auditable entity purges across all memory tiers with tombstones, fingerprints, and a structured purge log. - Use Case: After running a ranking check, ask the agent to update memory — it refreshes the hot cache hero keywords table, appends a rank-history snapshot, and flags significant movements for the next session. ## Quick Start Ask the agent to set up SEO memory for your project by providing your top keywords, primary competitors, and current metrics.

Frequently Asked Questions about memory-management

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

FAQPage Schema
How do I persist SEO project context across AI chat sessions?▼

Use a tiered memory structure: a hot cache file auto-loaded at session start, warm files loaded on demand, and cold archives for historical data. The skill handles promotion, demotion, and archival automatically based on reference frequency and staleness.

How do I delete personal data from project memory for GDPR compliance?▼

Invoke the purge command with the entity name. The skill greps all memory tiers including archives and wiki pages, deletes or anonymizes confirmed matches, writes a salted fingerprint tombstone to block reingest, and logs the purge with before/after grep counts for audit verification.

What is the hot cache line limit in this memory system?▼

The hot cache is capped at 80 lines and 25KB. When it exceeds these limits, the skill lists the oldest entries for archival and recommends trimming long entries during hygiene checks.

Can I retire old research files after they are compiled into wiki pages?▼

Yes, but only after passing five safety checks covering frontmatter capture, hash match, 90-day maturity, hot-cache absence, and compiled-page reference. Retirement is capped at 5 files per call and 20 per day, with full recovery via the originally_at frontmatter field.

Why does the wiki compile step fail on some systems?▼

Compilation requires the shasum tool to compute SHA-256 source hashes. On Windows or minimal containers where shasum is unavailable, the compile aborts with an explicit error rather than silently skipping hash verification.