pensyve
claudeUniversal memory runtime for Claude Code with cross-session memory, entity-aware recall, lifecycle hooks, skills, commands, and memory-curator agents
7 skills · 7 commands · 2 agents · hooks
Claude Code and Codex plugins their authors filed under "memory". All categories.
Universal memory runtime for Claude Code with cross-session memory, entity-aware recall, lifecycle hooks, skills, commands, and memory-curator agents
7 skills · 7 commands · 2 agents · hooks
Local session memory for coding agents: indexes existing session transcripts on disk across 43 coding agents into a local BM25 store with MCP recall, hooks, skill, and /deja command
2 skills · 1 command · 1 MCP server · hooks
Remembers approaches you tried and reverted (with reason and estimated token cost) and warns before you retry them, via a prompt-submit card and an ask-before-edit guard.
1 skill · hooks
KG state machine + Obsidian REST + auto-capture. Includes namespace aliases for chrono-kg / chrono-obsidian / chrono-catalog so legacy archive role plugins keep working.
3 skills · 3 MCP servers
Memory for your coding agent. Rekal saves every AI session transcript raw and append-only — no summaries to rot — then recalls the reasoning behind any change: why an approach was chosen, what was tried, what was rejected. Agent agnostic: Claude Code, Cursor, Codex, Gemini, Copilot, Kiro and OpenCode sessions all land in one ledger. Retrieval reasons at query time over the real record, with keyword and semantic search running locally — embedding is local, so there is no server, no account, no external memory service. Git is the transport: memory travels with your repo on push and reaches your team when work merges. Persistent cross-session context that maintains itself.
2 skills
Use Anamnesis as a local-first cross-agent memory bridge inside Claude Code.
1 skill · 2 commands · 1 MCP server
Persistent memory for Claude Code, built on Cognee's knowledge graph. Captures your prompts, tool calls and answers, recalls what's relevant on every prompt, and carries decisions and lessons into later sessions. Runs locally or on Cognee Cloud.
14 skills · 1 agent · hooks
Persistent causal memory for Claude Code — decisions, outcomes, and lessons that survive compaction. Same-context counterfactuals (natural experiments), a falsifiable prediction ledger, and an activation skill that teaches when to recall and record.
1 skill · 2 commands · 1 MCP server
Universal memory runtime — ambient working-memory substrate for Claude Code. Memory is not a feature you invoke — it's the substrate the agent operates on.
7 skills · 7 commands · 2 agents · hooks
AgentDB's own standalone MCP server (npx agentdb mcp start, pinned to the plugin version) with a memory specialist agent, vector-memory, reflexion-learning and setup-diagnostics skills, /agentdb-status, and a mod that keeps secrets out of memory and refuses destructive admin tools unless enabled. Supplies the agentdb server the other agentdb-* plugins call. Distinct from ruflo-agentdb, which bridges AgentDB through ruflo's MCP server.
3 skills · 1 command · 1 agent · 1 MCP server
Verifiable shared memory for AI agents: one address per place, one signature you check offline.
19 skills · 1 MCP server
Optional local signer for emem memory writes: your key stays in a sensitive option and one local process.
1 MCP server
Persistent, bitemporal memory for Claude Code: MCP tools, six skills, recall at session start and on every prompt, capture before compaction.
6 skills · 1 MCP server · hooks
Persistent memory system for AI coding sessions — cross-tool memory sharing with hybrid search
2 skills · 1 MCP server · hooks
Initialise la mémoire de session depuis le vault Obsidian — HOT_CACHE, patterns d'erreurs, daily log.
Persistent semantic memory. Slash command `/memo` + MCP tools `mcp__memo__memo_*`.
4 skills · 1 command · 1 MCP server · hooks
Project progress ledger written automatically. A Stop hook logs the working-tree delta each turn that touched files (ship, rfc, or plain conversation), so what moved is recorded without remembering to log; next reads it.
1 skill · hooks
Use GitHub or Linear issues as cross-session resolution memory. Create/update/resume self-contained, cold-readable issues, duplicates checked first.
1 skill
Session nap for Claude Code: /nap saves with trace, clears the session and resumes it with next by itself, under Herdr, herdr or tmux.
1 skill
Zero-config local memory for Claude, in Claude Code and Claude Cowork. Spawns a local demarkus-server and wires MCP tools for fetch/publish/append/graph/lookup over your own versioned markdown store, and self-documents sessions to it. /soul-join adds remote souls to a catalog with per-project binding and a destination gate; a join URL (host and token in one paste-able string) joins a remote soul in one step. With a promote destination configured (a joined knowledge system or a registered plain remote demarkus server), /promote lifts ready notes up to it, /promote-scan sweeps for candidates, an ADR publish nudges promotion, and /soul-refresh pulls promoted docs back as they evolve.
1 skill · 14 commands · 1 MCP server · hooks
Join an organizational demarkus knowledge system from Claude; a shared, broker-fronted, versioned markdown catalog. Steers sessions to the shared catalog, enforces publishing policy, and curates promoted notes. Runs no local server; a shared helper binary provides gates and guidance.
1 skill · 3 commands · hooks
A local MCP memory server: one SQLite file on your machine that Claude Code, Cursor, Windsurf, OpenAI Codex and Gemini CLI all read and write. No account, no credit card, no cloud, no API key.
1 skill · 1 MCP server
Memory that turns Claude into a counterpart. Local-first; everything stays on your computer.
4 commands · 1 MCP server · hooks
A pane beside the chat that shows what your counterpart's memory is doing, with a fox. Installs counterparts with it.
Durable, provenance-aware project memory for Claude Code.
5 skills · 1 MCP server · hooks
rein memory MCP server (40 tools, knowledge graph, adaptive engine, temporal recall, OAuth remote MCP)
1 MCP server
Persistent memory for Claude Code — store decisions, recall them semantically, and answer questions grounded in everything the agent has learned.
2 skills · 12 commands · 1 agent · hooks
Subconscious memory for Claude Code: recall before every turn, capture after every turn, backed by your own Redis, Valkey or PostgreSQL.
1 MCP server · hooks
Memory that abstains instead of guessing. Searches your project's own notes and decisions, marks a retracted claim as superseded, and refuses when nothing clears the threshold.
5 skills · 1 MCP server · hooks
Memory + recovery + code grounding. Restores deleted sessions, hybrid PG+Weaviate recall across every past conversation, tree-sitter symbol indexing across 12 languages, eight identifier-grounding domains (env / DB / route / GraphQL / flag / CSS / i18n / module-export), and a 14-strategy deterministic rewriter that catches fabrications before disk write. 89.3% reduction in hallucinated identifiers (51.5% → 5.5%, n=400).
33 commands · 1 MCP server · hooks