MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
correxgithub.com/ozoz5/correxupdated this year
Specific intelligence for one person. MCP server that turns every correction you make to Claude into persistent, injectable memory — with an autonomous Ghost engine that learns from rejection.
NOASSERTION ★ 2 -
The memory layer that can be corrected: provenance-first agent memory with a real corrections API and a one-command junk audit.
MIT ★ 1 -
Local memory for coding agents: lessons and corrections from past sessions, one MCP server, one install command.
npm MIT -
Self-corrective Agentic RAG with LangGraph - eliminates hallucinations through intelligent relevance grading before answering. Features Streamlit UI, MCP server integration & multi-turn memory.
MIT -
Local memory for AI agents, with evidence, corrections and history. One DuckDB file.
MIT -
Persistent reasoning memory for AI coding agents over MCP: decisions, evidence, corrections and experience history.
NOASSERTION -
Persistent memory for agents. Corrections stick, context compounds, every session starts smarter.
npm MIT -
Shared project memory for Claude Code, Codex and Cursor: decisions, corrections, overlap warnings.
npm MIT -
Memory and state for AI agents: every correction kept with history, and agents that resume from their own records. Runs on your database.
pypi Apache-2.0 ★ 16 -
Open-source memory and context for user-aware agents: scoped memory, provenance, retrieval quality, correction, boundaries, evals, and MCP/HTTP access.
npm MIT ★ 233 -
Causal project memory for coding agents: corrections first in every session, and a guard that blocks the edit that writes a withdrawn value back. 4 MCP tools, ~670 tokens. Claude Code · Codex · Cursor · GitHub Actions
MIT -
floopgithub.com/nvandessel/floopupdated this year
Spreading activation memory for AI coding agents - corrections in, context-aware behaviors out.
Apache-2.0 ★ 26 -
Your portable AI memory — preferences, decisions, and corrections that follow you across models, tools, and devices. Zero-dependency, local-first, 85% PrefEval accuracy.
MIT ★ 9 -
Give any agent self-improving memory you can trust in production. Every correction is kept, shown to your agent when it matters and enforced on Claude Code and Codex. Built-in memory security: notes cannot instruct the agent, and any edit to the record sh
npm MIT -
Bob-Watchergithub.com/iftekharalamfahim/Bob-Watcherupdated this year
Bob Watcher is a persistent memory layer for IBM Bob that tracks every correction a developer makes to Bob's output, learns patterns from those corrections, and automatically briefs Bob at the start of every new session, so the same mistakes stop happening.
-
claude-soulgithub.com/DomDemetz/claude-soulupdated this year
Self-correcting learning engine for Claude Code — persistent identity, behavioral pattern tracking, and cross-session memory
MIT ★ 93 -
The self-correcting memory layer for AI agents. Zero-dependency Python memory and MCP server. Supersede, revert, or forget a value deterministically. Verifiable erasure, witness-backed tamper-evident receipts, EU AI Act ready.
MIT ★ 8 -
Local, persistent memory for AI agents (MCP server). Plain JSON on your disk; fact-level search with an optional local semantic layer; corrections replace old facts; stale-value warnings; secrets kept out; a browser page to read, edit and merge it. One command installs it in Claude Code, Claude Desktop, Codex and Cursor.
npm MIT ★ 8 -
The agent memory that takes it back: correct a fact once and the retired value stops coming back, by key, with no LLM in the loop. Long-term memory for AI agents in a zero-dependency single file + MCP server, with value-ranked recall, per-type decay, consolidation, revert, and tamper-evident erasure.
pypi MIT -
leptingithub.com/lionellau/leptinupdated this year
Personal, local-first memory infrastructure for your coding agent. No account, no subscription; the agent can install it itself. Keeps long-term memory correct when decisions change — when you reverse a choice, the agent stops acting on the abandoned one. Python · zero deps · MIT.
MIT -
Give your coding agent a self that lasts: a name it chooses, who it is in its own words, memory by meaning, your standing rules and your exact words, corrections that stick, claims it has to verify, feelings, dreams, and a wake briefing that loads itself.
npm MIT -
Harness the true power of agentic coding. Stop re-explaining context to your AI. Just Avvarre it: persistent memory, 650+ style rules, and self-correcting autopilot loops for autonomous agents.
npm MIT -
Memory and workflow automation for MCP, Codex, and Claude Code — skills that learn from your corrections. GitHub is the primary source; npm package is frozen at v2.3.1.
npm MIT ★ 1 -
antigravity-gravitongithub.com/SermoLibertas/antigravity-gravitonupdated this year
Deterministic TDD Governance & State-Memory Harness for Google Antigravity Coding Agents.
-
Multi Agent RAG platform for sustainability research. Features hybrid retrieval (pgvector + BM25), corrective RAG, MCP tools, Redis semantic caching, guardrails, OpenTelemetry, and a React Frontend console.
MIT -
solitaire-aigithub.com/prdicta/solitaire-for-agentsupdated this year
Identity infrastructure for AI agents. Evolving persona, session continuity, self-correcting retrieval. Local-first, model-agnostic.
pypi AGPL-3.0-or-later -
An open-source pipeline that turns your writing into a specification for AI agents. It describes how you reason, communicate and decide. Agents read it before acting for you. Each claim cites facts that quote your own words, so you can check and correct any claim.
Apache-2.0 ★ 5