What is open-brain-server CLAUDE.md?

Project instructions for Open Brain, a self-hosted memory server that lets AI assistants save and search thoughts from Telegram or other clients.

They help contributors understand the capture and search flows, Supabase database structure, vector search, authentication, and the locations of shared code and edge functions.

They explain how captured thoughts are processed, stored, linked, and kept separate between users.

open-brain-server CLAUDE.md — Instructions file for Claude Code by Bobby-cell-commits: Project instructions for Open Brain, a self-hosted me

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add instructions/bobby-cell-commits/open-brain-server/claude-md

Clone the repo

git clone --depth 1 https://github.com/Bobby-cell-commits/open-brain-server

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/bobby-cell-commits/open-brain-server/claude-md.svg)](https://agentmods.dev/instructions/bobby-cell-commits/open-brain-server/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/bobby-cell-commits/open-brain-server/claude-md"><img src="https://agentmods.dev/badge/instructions/bobby-cell-commits/open-brain-server/claude-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 969 This file is loaded in full into every session.

When invoked 969 The same file — it is already loaded in full.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00969	$0.00969
Opus 5	$0.00485	$0.00485
Sonnet 5	$0.00194	$0.00194
Haiku 4.5	$0.00097	$0.00097

Measured 6d ago against content hash 1ca217c342b7, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

open-brain-server CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 70 lines

How it starts

The opening of the file, as written. The whole thing — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

Open Brain is a self-hostable MCP memory server for AI assistants. Thoughts are captured from Telegram (or any MCP client), embedded with vector search via OpenRouter, stored in Supabase with pgvector, and exposed through an MCP server so any AI assistant can search/write to the same shared memory.

Architecture

Capture path: Source → Edge Function → OpenRouter (embedding + metadata extraction in parallel) → Supabase insert → auto-link similar thoughts → extract entities

Retrieval path: MCP client call → Edge Function → tool execution → JSON response

Storage: Supabase Postgres + pgvector. thoughts table with 1536-dim embeddings, JSONB metadata, idempotency via (source, source_event_id) partial unique index. RPC match_thoughts for similarity search.

Auth: MCP key via x-brain-key header. Multi-tenant — each API key maps to an isolated brain.

Code Layout

supabase/functions/ — Deno-based Supabase Edge Functions
- _shared/ — Shared modules: supabase-client.ts, openrouter.ts, types.ts, errors.ts
- telegram-bot/ — Telegram webhook handler (primary capture path)
- open-brain-mcp/ — MCP server (14 tools). See docs/cookbook.md for usage guide.
- run-pipeline/ — Edge Function for automated RSS, HF Papers, and Emergent Mind ingestion
- monitor-pipeline/ — Pipeline health monitoring with Telegram alerting
supabase/migrations/ — Ordered SQL migrations applied via Supabase CLI
pipeline/ — Python-based pipeline scripts
- config.py — Env loading, subreddit list, feed URLs, focus terms
- openbrain_client.py — JSON-RPC client for capture_thought
- triage.py — LLM triage via OpenRouter
- dedup.py — JSON-file-based processed item tracker
- reddit/ — Subreddit monitoring (public .json endpoints)
- rss/ — RSS/Atom feed poller
- emergentmind/ — Emergent Mind arXiv trending papers
- hf_papers/ — HuggingFace daily papers
- briefing/ — Morning briefing generator
scripts/ — Bootstrap, deploy, validate scripts
tests/integration/ — Multi-tenant integration tests
docs/ — Cookbook, source-writing guide, tracker guide, skill showcases

Read the full file on GitHub · 70 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

6d ago First seen · 70 lines · 969 tokens per session scan A 1ca217c342b7

Subscribe to this mod's changes

open-brain-server CLAUDE.md is an instructions file published in the GitHub repository Bobby-cell-commits/open-brain-server (8 stars, last pushed 4mo ago), licensed MIT. It adds 969 tokens to every session, about $0.0048 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Memory and context

aeon CLAUDE.md

Claude Code instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).

aeonfun/aeon · 3,720 tokens

cognirepo CLAUDE.md

Claude Code instructions for ashlesh-t/cognirepo, covering claude.md, key rules, session start sequence (run in this order), behavioral confirmation rule and personas (cognirepo-402, cognirepo-403).

ashlesh-t/cognirepo · 2,241 tokens

wayland-core copilot-instructions.md

Copilot instructions for FerroxLabs/wayland-core, covering ijfw rules, output discipline, memory routing, context discipline and cross-audit.

FerroxLabs/wayland-core · 823 tokens

mcp-structured-memory CLAUDE.md

Claude Code instructions for nmeierpolys/mcp-structured-memory, a project described as: Structured Memory MCP Server.

nmeierpolys/mcp-structured-memory · 160 tokens

inkwell-memory CLAUDE.md

Instructions for veronchenko/inkwell-memory, covering claude.md — inkwellmemory, layout, multi-tenant mode (inkwellmultitenant=1), conventions and testing.

veronchenko/inkwell-memory · 2,378 tokens

RNR-Enhanced-Cognee AGENTS.md

AGENTS.md instructions for vincentspereira/RNR-Enhanced-Cognee, covering rnr enhanced cognee implementation for codex, critical requirements, 1. ascii-only output (no unicode encoding), 2. dynamic categories (no hardcoded categories) and 3. standard memory mcp interface.

vincentspereira/RNR-Enhanced-Cognee · 6,855 tokens