Persistent memoryfor Claude, Cursor& Copilot.

Stop re-explaining your project every session. ContextForge keeps Claude Code, Cursor, and Copilot in sync across every session via MCP — what you decided yesterday is still there today.

AI tools are powerful.But they have no memory.

You open Claude Code. You explain your architecture. Your naming conventions. The auth decision from last week. Again.

The more complex your project, the more time you waste giving context your agent should already know.

AI coding agents are powerful — but every conversation is a blank slate. That's the bottleneck.

Full terminal workflow with memory that carries across every session.

Persistent memory across conversations in the ChatGPT desktop app.

Rich conversations that remember your project, decisions, and context.

AI-assisted editing with context that persists between sessions.

Cascade AI with persistent project memory via MCP.

Agent mode in VS Code with persistent memory via MCP.

Add one JSON block to your MCP config. Your agent starts remembering — decisions, architecture, git history — everything.

"args": ["-y", "contextforge-mcp"],

"CONTEXTFORGE_API_KEY": "cf_sk_..."

Keep every repo organized. Group spaces, share with your team, and control who sees what.

Each project groups related spaces, issues, and collaborators. Your agent always knows which context belongs where.

📁 Tsubasa Soccer 10 spaces · 3 collaborators

Separate concerns cleanly. Each space has its own embeddings — your agent searches only what it needs, with zero noise.

Each space has its own embeddings index. Your agent searches only the context it needs — faster results, zero noise.

Ask in plain English. ContextForge understands meaning, not just keywords — so your agent always finds the right context.

Ask questions in plain English. ContextForge understands meaning, not just words — so your agent always finds the right context.

query: "how does authentication work?"

Keyword search fails when your code uses different words than your query. Semantic search understands what you mean — not just what you type.

Your code says "authentication", "auth",

1. JWT auth middleware with Supabase SSR

3. OAuth callback handler implementation

Same query. Same codebase. Only semantic search finds the answer.

Running several sessions at once? See who else is live on the project and what each is working on — in real time. Advisory, not a lock: never work blind next to your own agents again.

Presence never blocks you — it just makes sure you are never working blind. Every session sees what the others are focused on, right now.

✓ 2 other sessions live on "MCP-context-forge"

🟢 worktree-auth refactoring the auth module · 34s ago

🟢 tests-runner rebuilding the e2e suite · 1m ago

→ check 'focus' before touching the same areas

Commits, PRs, and branches sync automatically. Your agent always knows what changed — and why.

Connect once, stay updated forever. Every commit and PR is indexed automatically so your agent never falls behind.

Invite your team, manage roles, share knowledge. Every agent and every member — always in sync.

Your whole team shares the same knowledge base. Every agent, every member — always in sync.

✓ 3 collaborators on "Tsubasa Soccer"

Your agent creates tasks from code reviews, assigns priorities, and resolves them — all through MCP.

Your AI agent can create tasks from code reviews, assign priorities, and resolve them — all through MCP commands.

Save your agent's memory at any point. Accidentally broke something? Roll back instantly.

Accidentally corrupted your knowledge base? Roll back to any snapshot instantly. Your data is always safe.

Scoped keys for CI/CD, scripts, or custom tooling. Read-only or full access. Revoke anytime.

Create read-only or full-access keys per project. Revoke anytime. Perfect for CI/CD pipelines and automation.

Start free. Upgrade when you need more. 14-day free trial on paid plans.

Perfect for trying out ContextForge with your personal projects.

- 1 project
- 3 spaces
- 200 documents
- 500 semantic queries/mo
- 50 MB storage
- Batch operations
- Import / Export
- Git integration
- Snapshots
- Collaborators

For developers who need more power and team collaboration.

- 25 projects
- Unlimited spaces
- 5,000 documents
- 15,000 semantic queries/mo
- 1 GB storage
- Git integration — unlimited repos
- 50 snapshots
- 5 collaborators
- 2,000 webhooks/mo
- 14-day free trial

For teams and power users who need unlimited everything.

- Unlimited projects
- Unlimited spaces
- 10,000 documents
- 20,000 semantic queries/mo
- 2 GB storage
- Git integration — unlimited repos
- Unlimited snapshots
- Unlimited collaborators
- 5,000 webhooks/mo
- 14-day free trial

Your AI agent should know your project as well as you do. Give it memory that lasts.

This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.

Hebbian memory for AI agents — learns file access patterns, builds neural pathways, predicts next tools/files, saves tokens

ByteRover Team Memory is a paid remote MCP endpoint for OpenAI Codex memory MCP. It exposes structured JSON tools, a public server card, token-based access, usage receipts, and audit-ready w

Provides perfect memory for AI-assisted development by capturing project context snapshots, enabling natural language search, evolution tracking, and code intelligence.

Your AI's memory shouldn't live on someone else's server — 12 MCP tools that give it persistent context from your local markdown, no cloud, no API keys, single binary.

OpenMemBrain is the intelligent membrane for AI coding memory. It autonomously reads and learns from your coding sessions — you never have to tell it what to save. It selectively absorbs project knowledge, blocks secrets, filters noise, resolves conflicts, and persists only what matters.

Persistent AI memory across tools — remember your preferences, code standards, and decisions across Claude Code, Cursor, Codex, and any MCP tool. Local-first, zero-cloud.

Search and evaluate MCP servers from your AI agent: quality grades, live verification status, install commands and client compatibility for 5,000+ servers.

Interaction profile registry for AI agents — log interactions, build a behavioral profile, query it through behavioral lenses. 21 tools, zero-config.

Governed memory for coding agents with trust lifecycle, conflict detection, staleness tracking, and health scoring. SQLite + FTS5, zero infrastructure. Works with Claude Code, Cursor, Codex, Windsurf.

Persistent project memory + architectural decisions + pre-execution safety hooks for Claude Code. Local-only storage, multi-repo workspace, automatic knowledge extraction via background auditor.