More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Project Brain

    1

    One living knowledge graph for you and your AI agents: answers cite file and line, and you can browse it as a graph.

    $10

    /mo

    Subscription · cancel anytime

    Secure checkout via Stripe

    • Always the latest version
    • Delivered live to your agent via MCP
    • Cancel anytime, access ends at period end
    Security scanned
    project-brain

    project-brain

    Example session with this skill installed

    Set up the project brain and index my current documentation.

    • Read your context and instructions
    • Compiled the project-brain

    I have initialized the project brain, built the knowledge tool, and synchronized your documentation. The knowledge graph is now active and the task queue structure has been created.

    Knowledge Setup Report

    ComponentStatusDetails
    Knowledge GraphActive12 docs, 84 sections indexed
    MCP ServerConfiguredAdded to .mcp.json
    Task QueueReadyFolders created in tasks/
    Agent FileCreatedCLAUDE.md initialized with project rules

    Knowledge Statistics:

    • Domains: architecture, api-spec, internal-docs
    • Pending: 5 nodes need summaries.
    • Storage: knowledge_data/knowledge.db (gitignored)

    Next steps

    1. Run K view to open the knowledge graph in your browser at http://127.0.0.1:7474.
    2. Decide if you want to generate the 5 pending summaries using the AI step or manual annotation.
    3. Add your first task to the tasks/ folder to begin the orchestrated workflow.
    $10/mo

    Subscription · always the latest version

    • Secure checkout via Stripe
    • Cancel anytime
    • Delivered live via MCP

    About this skill

    Your project's shared brain, for you and your AI agents.

    Agents forget everything between sessions, and your project's knowledge is scattered across docs, notes, specs and finished tasks. project-brain turns every markdown file into one living knowledge graph that your agents query before they answer, and that you can browse in your browser.

    What you get

    • Answers you can check: every answer cites the file and line it came from.
    • A graph you can see: a local web view where domains open into docs and sections, with summaries, links, search and filters.
    • Knowledge that knows when it's outdated: changed text is flagged for a new summary, and summaries you wrote are never overwritten.
    • Work that becomes knowledge: a task queue that parallel agent sessions work through, where each finished task records what changed and how it was checked, linked into the graph.
    • An agent that learns: your corrections become dated rules, in your own words.
    • Web pages as sources: keep any page as project knowledge, with its URL.

    How it works Your markdown files stay the source of truth. san_knowledge, the engine (Go, an embedded graph database, an MCP server and the web view), ships as readable source and is built once on your machine. It runs locally, with no API key and no vector database; optional AI summaries use your own Claude Code login.

    Requirements: Go 1.25+, Node.js 20+ (for the graph view). Made for Claude Code; the MCP server works with any MCP-capable agent.

    Requires Go 1.25+ and Node 20+. Indexes markdown only. Graph view requires local browser access. Summaries for large projects may incur token costs if processed via AI.

    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 28 days ago

    Requires

    GitHub
    X
    Datareportal
    Go
    Node.js

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Network Access

    Allowed Hosts

    example.com
    datareportal.com
    github.com
    www.bps.go.id
    x.com
    Listed28 days ago

    What's inside

    Frequently Asked Questions

    Project Brain solves 'agent amnesia' by creating a persistent knowledge graph and task queue that lives outside the chat window, allowing different AI sessions to share the same context, rules, and progress.

    This skill works with any AI agent or environment that supports the Model Context Protocol (MCP), such as Claude Desktop, Windsurf, or Cursor, by exposing specialized tools for knowledge retrieval.

    The skill indexes all Markdown files in your project root to build relationship links and summaries, but it uses those docs as a map to provide precise file-and-line citations for your source code.

    Your purchase includes the full source code for the Go-based knowledge engine, the React visualization UI, the MCP server integration, and the CLI tools for manual graph management.

    Setup requires Go (1.25+) and Node (20+) on your host machine to compile the 'san_knowledge' binary and build the UI, after which you simply point your agent to the generated MCP configuration.

    Unlike manual prompting, this skill uses a structured 'bbolt' database and a file-based state machine, ensuring project rules and task statuses remain consistent even when you hit LLM context limits.

    New in Agents & Orchestration