skills.rest
abdulazeezoj
abdulazeezojCOMMUNITY
Explore all skills by this author →
Website𝕏@abdulazeezojGitHub

setup-gbrain

Installs and configures the gbrain knowledge base CLI and MCP server for coding agents.

Updated Sep 13, 2026
Search Tags:#claude code#knowledge base#supabase#mcp#setup#pglite#gbrain
One-click install
npx skills add https://github.com/abdulazeezoj/monovella-poc --skill setup-gbrain-abdulazeezoj
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: setup-gbrain
Source: https://github.com/abdulazeezoj/monovella-poc/tree/main/.agents/skills/gstack/setup-gbrain
Command: npx skills add https://github.com/abdulazeezoj/monovella-poc --skill setup-gbrain-abdulazeezoj

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Setting up a persistent, searchable knowledge base for a coding agent involves many fragile steps: installing the gbrain CLI, choosing and initializing a storage engine (local PGLite, Supabase Postgres, or a remote MCP server), registering MCP tools, and capturing per-repo trust policies. This Skill automates that entire onboarding flow with idempotent, rollback-safe steps. ## Core Features & Use Cases - Multi-path engine setup: Initialize gbrain against an existing Supabase URL, auto-provision a new Supabase project, run a local PGLite database, or connect to a remote gbrain MCP server over HTTP with a bearer token. - Broken-engine remediation: Detects dead or misconfigured engines and offers retry, rollback-safe switch to PGLite, or full re-initialization. - Transcript and memory ingest: Optionally ingests Claude Code, Codex CLI, and Cursor session transcripts plus curated gstack artifacts into gbrain, with per-remote trust policies (deny/read-only/read-write) and optional gitleaks secret scanning. - Use Case: A developer on a new machine runs the setup once and gets gbrain installed, MCP-registered in Claude Code, transcripts indexed, and a GBrain Configuration block persisted to CLAUDE.md so future sessions auto-load prior context. ## Quick Start Ask the agent to set up gbrain on this machine, for example: "Run /setup-gbrain to install gbrain, initialize a local PGLite brain, and register it as an MCP server."

Frequently Asked Questions about setup-gbrain

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up gbrain for Claude Code?▼

Run the /setup-gbrain skill, which installs the gbrain CLI, initializes a storage engine, and registers it via claude mcp add at user scope. The flow is idempotent, so re-running it repairs missing pieces without duplicating state.

What storage engines does gbrain support?▼

gbrain supports a local PGLite database, a Supabase Postgres backend via the Session Pooler URL, and a remote MCP server over HTTP with a bearer token. The setup skill offers all paths, including auto-provisioning a new Supabase project.

Can I use gbrain on a remote server instead of locally?▼

Yes, Path 4 registers a remote gbrain MCP endpoint over HTTPS with a bearer token, verified before registration. You can optionally add a local PGLite engine just for symbol-aware code search while transcripts route to the remote brain.

What happens if my gbrain engine is broken or unreachable?▼

The skill detects broken-db or broken-config states and offers a remediation prompt: retry the probe, switch to local PGLite with a rollback-safe config backup, re-initialize from scratch, or quit. Failed PGLite inits restore the previous config automatically.

Does gbrain transcript ingest scan for secrets?▼

Per-file gitleaks scanning during memory ingest is opt-in via a flag or environment variable and is off by default. Sessions with positive findings are skipped entirely, and per-remote deny or read-only trust policies block ingestion before any writes.

Where is the gbrain bearer token stored?▼

The bearer token for remote MCP mode is stored in ~/.claude.json with mode 0600, where Claude Code keeps all MCP credentials. It is never written to CLAUDE.md, since that file is often committed to git.

Related Skills

Explore More Software Engineering Skills→
tblakex01tblakex01
1

setup-gbrain

Install the gbrain CLI, initialize a local brain, and register MCP.

Community
Advanced
cloudofgeorgecloudofgeorge

sync-gbrain

Synchronize GBrain data with the repository and refresh CLAUDE.md search guidance.

Community
Advanced
harsh-pandheharsh-pandhe

setup-gbrain

Install and configure gbrain with MCP registration for Claude Code on Mac.

Community
Advanced
davpatel605-beepdavpatel605-beep

chroma

Store and retrieve vector embeddings for semantic search and document retrieval.

Community
Intermediate