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.
npx agentmods add instructions/ivncmp/dbrain/claude-mdgit clone --depth 1 https://github.com/ivncmp/dbrainWrote 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.
[](https://agentmods.dev/instructions/ivncmp/dbrain/claude-md)<a href="https://agentmods.dev/instructions/ivncmp/dbrain/claude-md"><img src="https://agentmods.dev/badge/instructions/ivncmp/dbrain/claude-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.03935 | $0.03935 |
| Opus 5 | $0.01968 | $0.01968 |
| Sonnet 5 | $0.00787 | $0.00787 |
| Haiku 4.5 | $0.00394 | $0.00394 |
Grade C, and why
dbrain CLAUDE.md scanned grade C with 3 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 3d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
- `~/.claude/settings.json` — Tool permissions Unrestricted tool accessmediumExcessive agency
A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.
- **Permissions in `~/.claude/settings.json`** — Use `"mcp__dbrain__*"` wildcard to allow all tools globally without per-project prompts. Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -H "Authorization: Bearer sk-dbr_..." \ How it starts
The opening of the file, as written. The whole thing — 384 lines — stays where its author put it; the contents beside it link to each section on GitHub.
dbrain
Your distributed mind. Wherever you go, I remember.
What is it
dbrain is a brain for your AIs. It stores who they are, who you are, and everything you've done together. Install it once, and every AI you use — at home, at work, on mobile — connects to the same brain. Same identity, same memories, same knowledge. No matter where you are.
[Home] Claude Code ──MCP──┐
[Work] Claude Code ──MCP──┤ ┌───────────────────────┐
[Mobile] Gemini ──REST──────┼────→│ dbrain (your mind) │
[Server] OpenClaw ──REST────┤ │ identity + memory + knowledge
[Other] Custom AI ──API────┘ └───────────────────────┘
Architecture
The brain has 4 layers:
- Identity — Who am I? Who are you? How should I behave? Stored in
documents. Every AI reads this first when it connects. One identity, many machines. - Conversations — Raw chat history from every AI session. Every message pair (user <-> assistant) gets stored. The brain remembers every conversation, from every machine.
- Knowledge — Structured facts organized by PARA (Projects/Areas/Resources/Archives). Typed entities (person, project, event, system). Hot/warm/cold tiers by access frequency. Memories that matter stay hot, the rest fades — like a real brain.
- Recall — Full-text search (FTS5) over all facts. Ask a question, get the memory.
Optional: with an LLM configured, dbrain periodically reviews unprocessed conversations and extracts facts into the knowledge layer. Like a brain consolidating memories during sleep.
Stack
| Layer | Technology |
|---|---|
| Language | Node.js + TypeScript |
| API | Fastify |
| DB | SQLite + FTS5 (better-sqlite3) |
| MCP | @modelcontextprotocol/sdk (HTTP transport) |
| Validation | Zod |
| CLI | @clack/prompts (interactive wizard) |
| Dashboard | Single-file React app (CDN, no build step) |
| Port | :7878 (REST API + MCP HTTP, same port), :7879 (Dashboard) |
Installation & Usage
Two-step process: init the brain (server), then connect clients.
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.
- 3d ago First seen · 384 lines · 3,935 tokens per session scan C 5b4dda9d0b35
dbrain CLAUDE.md is an instructions file published in the GitHub repository ivncmp/dbrain (3 stars, last pushed 3mo ago), licensed MIT. It adds 3,935 tokens to every session, about $0.0197 per session on Opus 5. A static security scan graded it C with 3 findings (reads agent configuration directories, unrestricted tool access, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
Recall AGENTS.md
AGENTS.md instructions for samzong/Recall, covering agents.md, commands, architecture and boundaries.
opencode-claude-memory AGENTS.md
Instructions for kuitos/opencode-claude-memory, covering agents.md, structure, where to look, critical coupling and conventions.
vibepod-cli AGENTS.md
AGENTS.md instructions for VibePod/vibepod-cli, covering project agent memory, adding a supported agent, tests and maintaining this file.
jaz AGENTS.md
AGENTS.md instructions for gluonfield/jaz, covering engineering rules, backend architecture and integrations goal.
obsidianos_work AGENTS.md
Instructions for benoror/obsidianos_work, covering agents, skills, skill proxies (symlinks to .agents/skills/), rules and agent wrappers.
DCGP AGENTS.md
Instructions for addicted2crypto/DCGP, covering dcgp - agent mounting guide, conformance tier claimed by this repo, what this system does, the 7-step circular intelligence loop and the entropymonitor formula.