Cross-Agent Memory Bridge

    1

    Free, MIT-licensed local-first AI memory skill. Manage portable, human-approved context snapshots with privacy checks, controlled exports and rollback. Python 3.11+; no automatic cross-platform sync.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Cross-Agent Memory Bridge

    Example session with this skill installed

    Initialize my local vault and show me a summary of the current facts you have on file for my developer profile.

    • Read your context and instructions
    • Compiled the cross-agent memory bridge

    I have successfully initialized and validated your local memory vault. The integrity check confirms all files are consistent with the latest recorded hashes.

    Vault Initialization Receipt

    FieldValue
    Version1.2.4
    Content SHA256e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855
    Source/users/dev/vault/START_HERE.md
    Loaded Filesprofile.json, preferences.yaml, history.log
    StatusVERIFIED
    Loaded At2024-05-20T14:30:05Z

    Current Profile Summary

    • Primary Stack: TypeScript, Node.js, PostgreSQL.
    • Architecture Preference: Hexagonal, Domain-Driven Design.
    • Documentation Style: Technical, concise, Markdown-first.
    • Build Tools: Turborepo, pnpm.

    Next steps

    1. Run propose --update if you need to modify your primary stack or tool preferences.
    2. Use export --recipient "Writing Agent" to prepare a sanitized background for a new session.
    3. Perform verify-snapshot if you manually edited any vault files outside of this bridge.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Cross-Agent Memory Bridge — Free & Open Source

    Keep useful context portable across AI agents without treating every remembered detail as automatically shareable.

    Cross-Agent Memory Bridge is a local-first, MIT-licensed Agent Skill and Python CLI for managing structured personal/project context that you explicitly approve. It focuses on reviewability, privacy and version history rather than a hosted memory service.

    What it provides

    • Local context vaults with clear record status, source, date and version information
    • Proposed edits, visible diffs and explicit human approval before applying confirmed changes
    • Privacy checks and preview-before-export for shareable snapshots
    • Snapshot hash and freshness checks, plus auditable rollback
    • No hosted database, telemetry, subscription or paid API required for core CLI operations

    Requirements and limitations

    Python 3.11 or later. Git is optional. The core runs locally. There is no automatic perpetual sync between ChatGPT, Claude, Gemini or other agents, and no claim that all integrations are natively tested. Manual Markdown import/refresh may be required.

    Agensi's MCP delivery is a discovery/access channel, not an automatic installation of the complete local Python workflow. For full CLI use, review the source and follow the installation guide on GitHub. Keep your personal vault outside the Skill's installation directory.

    Open source: https://github.com/rambolanbohu-bot/cross-agent-memory-bridge

    If you find it useful, consider starring the GitHub repository or opening an issue with feedback.

    Requires local vault setup. No automatic remote uploads. Excludes health, financial, and auth data. Python 3.11+ required.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 8 days ago

    • Free with an Agensi account

    Requires

    GitHub
    Python

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Network Access

    Allowed Hosts

    api.github.com
    Listed8 days ago

    What's inside

    Frequently Asked Questions

    This skill creates a secure, human-verified bridge for your personal context, preventing different agents from hallucinating or forgetting your preferences, professional background, and project details. It uses SHA-256 hashing and mandatory approval steps to ensure your data remains accurate and private as you move between sessions.

    The skill is designed for Python 3.11+ environments and is platform-agnostic, meaning it can interface with any AI agent that can execute Python scripts or read local files. It operates primarily offline using your local filesystem, ensuring compatibility without requiring specific API integrations.

    Privacy is maintained through a "human-in-the-loop" filtering system and local-first storage. Before any data is exported to a new agent, the skill performs a privacy scrub and generates a preview, allowing you to manually approve exactly which facts are shared.

    The package includes the core Python logic for vault validation, change management, and privacy filtering, along with documentation for setting up your local vault. You also receive the integrity verification scripts and recovery mechanisms for managing data rollbacks.

    No, this skill is strictly designed for local filesystem operations and does not support automatic cloud uploads or Git remote actions. This architecture is intentional to ensure your sensitive context profiles never leave your local machine without your explicit consent.

    New in Agents & Orchestration