cognitivemyriad avatar

Verified Repo Memory

cognitivemyriad/mcp-verified-repo-memory

Gives Claude a verifiable memory system scoped to a single repository. You store facts with file citations (path and line range), and when retrieving them later, the server rehashes the actual code on disk to verify nothing changed. If the snippet moved, it relocates the citation automatically. If it changed or got deleted, the memory is marked stale and filtered out. Exposes five tools: vrm_store for saving memories, vrm_search for keyword lookup, vrm_retrieve for JIT-verified results, vrm_list for auditing by status, and vrm_forget for deletion. Memories live in a local JSON file at the repo root with a default 28-day TTL. Useful when you want an agent to remember refactoring decisions or codebase conventions without serving outdated information after the code evolves.

GitHub stars
1
tools
5
transport
STDIO
env vars
4
last updated
Feb 21, 2026

Tools

5 tools

Public tool metadata for what this MCP can expose to an agent.

vrm_storeStore a new memory with file citations.

Store a new memory with file citations.

No parameter schema in public metadata yet.

vrm_searchSearch for candidate memories by keywords.

Search for candidate memories by keywords.

No parameter schema in public metadata yet.

vrm_retrieveJIT-verify candidates and return only valid memories. (Main tool for agents)

JIT-verify candidates and return only valid memories. (Main tool for agents)

No parameter schema in public metadata yet.

vrm_listList memories by status (valid, stale, missing).

List memories by status (valid, stale, missing).

No parameter schema in public metadata yet.

vrm_forgetManually delete a memory.

Manually delete a memory.

No parameter schema in public metadata yet.

Sponsors

Configuration

VRM_TTL_DAYS

Default TTL in days (1-365).

VRM_DATA_DIR

Override data dir path.

VRM_MAX_FILE_BYTES

Max file parsing size bytes.

VRM_SECRET_SCAN

Enable secret scan (on/off).

Readme

Verified Repo Memory Banner

Verified Repo Memory MCP v0.1.2

CI TypeScript License: MIT

Stale-proof repository memory with citations + just-in-time verification + TTL (repo-scoped).

An MCP server providing "safe memory" for AI coding agents. Memories are scoped per repository, backed by code citations, and verified just-in-time so an agent never receives stale information when the underlying code has changed.

Quickstart

Run via npx:

npx -y @cognitivemyriad/vrm-local --repo /path/to/repo

(Alternatively, run from source: npm ci && npm run build && node build/index.js --repo /path/to/repo)

Tools


View the full README on GitHub

Details

Repository
github.com/cognitivemyriad/mcp-verified-repo-memory
Category
AI & LLM Tools
Package
@cognitivemyriad/vrm-local
Transport
STDIO
Registry
MCP registry, active

Related AI & LLM Tools MCP servers

View all
1dingdawg avatarDingdawg Legal Agentdingdawg/dingdawg-agent-1AI contract review, clause drafting, compliance checklists & legal research.0stars
2dingdawg avatarDingdawg Support Agentdingdawg/dingdawg-agent-1AI support agent — ticket classification, response drafting, escalation routing.0stars
3sharedmemoryai avatarMemorysharedmemoryai/mcp-serverPersistent memory for AI agents. Remember, recall, and explore a shared knowledge graph.0stars
4velixarai avatarMemoryvelixarai/velixar-mcp-serverPersistent memory for AI assistants. Store, search, and recall across sessions.1stars
5kaaustubh avatarProject Memory Mcpkaaustubh/project-memory-mcpPersistent memory for AI coding agents — past bugs, decisions, and your corrections, in your repo.1stars
6romainsantoli-web avatarMemory OS AIromainsantoli-web/memory-os-aiAdaptive memory for AI agents — FAISS search, chat extraction, cross-project linking1stars