skills.rest
Synaptic-Labs-AI
Synaptic-Labs-AIOFFICIAL
Explore all skills by this author →
WebsiteGitHub

pact-memory

Store and retrieve structured memories in a SQLite-backed store for PACT agents.

71|14|Updated Jun 26, 2025
Search Tags:#claude#memory#graph#embedding#semantic-search#pact
One-click install
npx skills add https://github.com/Synaptic-Labs-AI/PACT-Plugin --skill pact-memory-synaptic-labs-ai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: pact-memory
Source: https://github.com/Synaptic-Labs-AI/PACT-Plugin/tree/main/pact-plugin/skills/pact-memory
Command: npx skills add https://github.com/Synaptic-Labs-AI/PACT-Plugin --skill pact-memory-synaptic-labs-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires pysqlite3, sqlite-vec, model2vec, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Persistent memory stores context, goals, lessons, decisions, and entities across sessions and enables semantic search for cross-session learning.

Core Features & Use Cases

  • Rich memory objects: context, goals, tasks, lessons, decisions, and entities
  • Semantic search and graph-enhanced retrieval across memories and files
  • CLAUDE.md integration: sync working memory and retrieved context to support rapid agent recall

Quick Start

Save a memory after completing a PACT phase to capture context, goals, and lessons learned.

Frequently Asked Questions about pact-memory

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

FAQPage Schema
How do I persist agent memory across sessions for semantic search?▼

Agent cross-session memory persistence is achieved by storing structured context, goals, and lessons in a SQLite-backed memory store. This enables cross-session learning and semantic search retrieval for subsequent interactions.

What is graph-enhanced retrieval for agent memory and how does it work?▼

Graph-enhanced retrieval maps file relationships across projects to enrich semantic search results. It leverages a SQLite-backed store with an optional vector index to query structured memories and interconnected project files.

Does sqlite-vec work with model2vec for embedding generation in memory stores?▼

Yes, sqlite-vec works with model2vec to provide the optional vector index and embedding backend. This combination allows a SQLite-backed memory store to perform semantic search on structured agent memories.

How do I sync working memory and retrieved context to CLAUDE.md?▼

You can sync working memory and retrieved context to CLAUDE.md using built-in synchronization features. This updates the file to support rapid agent recall of structured memories captured during previous sessions.

What's the best way to capture lessons learned after completing a PACT phase?▼

The best way to capture lessons learned after a PACT phase is to save a structured memory object. This stores context, goals, tasks, decisions, and entities into the persistent SQLite-backed memory store for later retrieval.

Related Skills

Explore More Software Engineering Skills→
spqianspqian
3

dream

Consolidate agent memories into a three-tier graph with vector indexing.

Community
Advanced
severity1severity1
154

memory-processor

Update AUTO-MANAGED sections in CLAUDE.md files after code changes.

Community
Basic
AngelMaldonadoAngelMaldonado

cx-memory

Store and organize project memories in a SQLite-backed memory store.

Community
Intermediate
MichielDeanMichielDean

llmem

Store and retrieve LLM memories in SQLite with semantic search.

Community
Advanced