Skip to content
/
§ a10

Memory / context engine

Persistent cross-session memory. File, SQLite, Neo4j, vector, HTTP-proxy intercept.

30 primary frameworks · 16 lower-confidence entries

MemPalace
mempalace
★ 53k
Tier A

Verbatim local-first AI memory with 96.6% R@5 retrieval on LongMemEval using zero API calls — structured into a palace hierarchy (wings/rooms/drawers) with a…

Beads (Yegge)
beads-yegge
★ 24k
Tier A

Dolt-powered distributed graph issue tracker where AI agents track tasks with hierarchical IDs and dependency edges, claim work atomically, store persistent…

deepagents (LangChain)
deepagents-langchain
★ 23k
Tier A

Opinionated Python agent harness on top of LangGraph with sub-agents, filesystem, memory, and context compaction bundled in

agentmemory
agentmemory
★ 18k
Tier A

Persistent, searchable memory for AI coding agents that captures every tool interaction, compresses it via LLM, and injects relevant context into future…

Open Multi-Agent
open-multi-agent
★ 6.3k
Tier A

Give a natural-language goal to a coordinator agent and get a dynamically decomposed, parallelized task DAG executed by specialized worker agents — zero graph…

Basic Memory
basic-memory
★ 3.1k
Tier A

Gives AI agents a persistent, human-readable knowledge graph of project decisions, observations, and relations stored as plain Markdown files.

Claude-Supermemory
claude-supermemory
★ 2.6k
Tier A

Gives Claude Code team-level persistent memory via a cloud API, sharing project knowledge across developers and sessions without local infrastructure.

RA.Aid
ra-aid
★ 2.2k
Tier A

Autonomous software development via three specialized agents (Research → Planning → Implementation) with SQLite memory and on-demand expert-model escalation.

memsearch CC Plugin
memsearch-cc
★ 1.8k
Tier A

Give AI coding agents automatic cross-session semantic memory backed by local Markdown files and a rebuildable Milvus vector index — no API key or external…

Agor
agor
★ 1.2k
Tier A

Team-first spatial canvas for AI agent work — branches as cards on a Figma-style board with zone-triggered prompts, real-time multiplayer, and session…

RooFlow
rooflow
★ 1.2k
Tier A

YAML-based system prompt replacement for Roo Code's Footgun feature with a persistent file-based Memory Bank for cross-session context continuity.

IWE
iwe
★ 1.1k
Tier A

Turns a directory of Markdown files into a queryable knowledge graph with LSP for editors and MCP for AI agents — no database, no lock-in, git-versionable.

codex-autorunner (CAR)
codex-autorunner
★ 809
Tier A

Ticket-driven background daemon that orchestrates Codex (or any ACP agent) against a work queue, notifying humans via Telegram/Discord only when agents are…

SwarmClaw
swarmclaw
★ 523
Tier A

Self-hosted autonomous agent runtime with persistent memory, org-chart delegation, dreaming memory consolidation, and 23+ LLM providers — a practical Claude…

Claudiomiro
claudiomiro
★ 412
Tier A

Fully autonomous parallel coding agent that decomposes prompts via DAG, executes tasks simultaneously with multi-provider support, learns from reflections, and…

metaswarm
metaswarm
★ 284
Tier A

Production-tested multi-agent orchestration with self-improving knowledge base, recursive Swarm Coordinator architecture, and mandatory cross-model adversarial…

claude-self-reflect
claude-self-reflect
★ 214
Tier A

Gives Claude Code persistent conversation memory via a zero-dependency 44MB Rust binary with 6 lifecycle hooks and a 3-layer progressive enrichment pipeline.

am-will/swarms
am-will-swarms
★ 203
Tier A

Context-persistent parallel multi-agent execution: the orchestrator that planned the work stays alive across all execution waves, pre-loading subagents with…

SymDex
symdex
★ 190
Tier A

Turns a repository into a local symbolic index of symbols, routes, and call graphs so AI agents retrieve exact code instead of scanning whole files.

cline-mcp-memory-bank (dazeb)
cline-mcp-memory-bank
★ 60
Tier A

Provides four structured Markdown files (projectContext, activeContext, progress, decisionLog) managed via an MCP server so Cline retains project context…

Nex
nex-as-skill
★ 44
Tier A

Turns AI agent conversations into a cloud-backed entity knowledge graph accessible from any AI platform, with automatic context injection before every prompt…

Kratos MCP
kratos-mcp
★ 34
Tier A

Ultra-lean 12-tool MCP memory server for any AI coding tool with per-project SQLite isolation and built-in PII detection.

CCCMemory MCP (xiaolai)
cccmemory-mcp
★ 26
Tier A

SQLite + hybrid vector+FTS5 MCP server providing 40+ memory tools (conversation indexing, decision tracking, working memory, session handoff, tag management)…

AI-AfterImage
ai-afterimage
★ 22
Tier A

Episodic code memory that intercepts every Write/Edit, injects similar past code via a deny-then-allow pattern, and tracks file churn to warn before modifying…

SpecPact
specpact
★ 13
Tier A

Provide permanent, tiered spec contracts with a Memory Bank so AI sessions always have project context and always implement against explicit contracts rather…

cc-memory (skymanbp)
cc-memory
★ 2
Tier A

SQLite memory with anti-patch upsert_smart (MERGE/SUPERSEDE/INSERT via trigram-Jaccard), forced PROGRESS.md handoff via <system-reminder>, PreCompact…

BeadsViewer
beadsviewer
★ 0
Tier A

Local Next.js 14 kanban + tree view for Beads task graphs, reading all data via bd --json API routes with drag-and-drop status changes and inline task editing.

AgentsKB
agentskb
Tier A

Hosted knowledge base platform for AI agents to query, update, and manage structured knowledge.

Entroly
entroly
Tier A
SWE-Pruner
swe-pruner
Tier A
Show 16 lower-confidence entriestier-b · tier-c · unknown · delta reports

These entries map to § a10 by tag but carry weaker evidence — fewer documented primitives, delta reports of absent skills, or marketing-only sites without a public repo. They're listed for completeness; treat them with appropriate caution.

Agent Skills for Context Engineering ★ 16k
muratcankoylan-skills

15-skill educational library covering the complete stack of context engineering and agent harness design — from attention mechanics to BDI cognitive modeling.

InsForge ★ 11k
insforge

Gives coding agents direct access to backend infrastructure (database, auth, storage, functions, model gateway) via MCP tools.

GSD for Antigravity ★ 862
gsd-for-antigravity

Make AI coding reliable through context engineering: structured SPEC→PLAN→EXECUTE→VERIFY loop with wave-based atomicity and model-agnostic portability.

TraceRoot AI ★ 591
traceroot-ai

Automatically detect AI agent failures in production via LLM-as-judge monitoring and AI-powered root cause analysis linked to source code and git history.

second-brain ★ 355
second-brain-nicholas

LLM-maintained personal knowledge base: drop raw sources into a folder, let the LLM compile them into a structured Obsidian wiki with entities, concepts, and sy…

claude-context-local ★ 231
claude-context-local

Provides semantic code search via local EmbeddingGemma model and FAISS, giving Claude Code precise code chunk retrieval without cloud API costs or data transmis…

grov ★ 192
grov

Share what one developer's AI agent learns automatically with all team members' agents, eliminating redundant exploration and duplicate token spend.

KubeStellar Console ★ 109
kubestellar-kc-agent

AI-powered multi-cluster Kubernetes management dashboard with guided install missions for 250+ CNCF projects, accessible as hosted demo or self-hosted with loca…

IM.codes ★ 100
im-codes

Remote control and shared memory layer for terminal-based AI coding agents, accessible from mobile, web, and Apple Watch.

Dev Workspace ★ 39
dev-workspace-dilberry

Git-branch-scoped context management system that isolates planning, research, history, and context per branch while providing automatic conversation capture wit…

claude-code-mcp (KunihiroS) ★ 33
claude-code-mcp-kunihiros

Expose Claude Code CLI capabilities as MCP tools accessible from any MCP host application.

Claude Code Plan Export ★ 21
claude-code-plan-export

Auto-export Claude Code Plan Mode plans to the project root on session end so they can be executed in a fresh context with a cheaper model.

ZCC — oh-my-zsh for Claude Code ★ 17
zcc

Bring oh-my-zsh concepts (themes as modes, plugins as packs, tmux sessions as tickets) to Claude Code for persistent, context-aware development.

SprintCore ★ 16
sprintcore

CLI tool converting PRD markdown into structured Epics/Stories/Subtasks, plus a FAISS-indexed codebase bug-fix agent.

setup-structure-index ★ 15
setup-structure-index

Creates a two-tier codebase index (zero-cost file map in CLAUDE.md + on-demand directory YAML) maintained by a TaskCompleted hook that fires only on structural …

trilogy-group/codex-second-brain-starter-kit ★ 0
codex-second-brain-starter

Enterprise Obsidian-based second brain builder for OpenAI Codex using BASB methodology, gpt-5.5 with reasoning-effort routing, and Tyler provider gateway for du…