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/pintomatic/kernal/claude-mdgit clone --depth 1 https://github.com/pintomatic/kernalWhat 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 | $0.02077 | $0.02077 |
| Opus 5 | $0.01038 | $0.01038 |
| Sonnet 5 | $0.00415 | $0.00415 |
| Haiku 4.5 | $0.00208 | $0.00208 |
Grade A, and why
kernal 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 2d 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.
How it starts
The opening of the file, as written. The whole thing — 188 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Kernal — Developer Guide
What This Is
Kernal is an open-source MCP server that provides a structured knowledge graph for professionals. It stores people, organizations, activities, topics, actions, and notes in SQLite, connected by a generic relationship graph. Claude (or any LLM) drives extraction — the server is a clean data store, the LLM is the brain.
Repo: github.com/pintomatic/kernal
License: MIT
npm: @kernal/mcp (not yet published)
Architecture
User ──→ Claude ──→ MCP (stdio or HTTP) ──→ Kernal Server ──→ SQLite
↑
Dashboard ──→ REST API ──┘
- Local mode: Stdio transport, Claude Desktop connects directly
- Cloud mode: Express.js with StreamableHTTP MCP transport + REST API
- Dashboard: Vite + React app, connects to cloud server REST endpoints
Stack
- TypeScript 5.8, Node 18+
- SQLite via better-sqlite3 (local) / @libsql/client (Turso, cloud-ready)
- MCP SDK v1.27.1 (@modelcontextprotocol/sdk)
- Express.js 5 (cloud server)
- Zod (input validation)
- React 19 + Vite + Tailwind CSS 4 + react-force-graph-2d (dashboard)
- Vitest (testing)
Project Structure
kernal/
├── src/
│ ├── index.ts # MCP stdio entry point
│ ├── server.ts # McpServer + tool registration
│ ├── cloud.ts # Express.js cloud entry point (HTTP + REST + MCP)
│ ├── db/
│ │ ├── schema.sql # DDL (7 tables + FTS5 + triggers)
│ │ ├── adapter.ts # DbAdapter interface
│ │ ├── local.ts # better-sqlite3 adapter
│ │ └── turso.ts # @libsql/client adapter (async)
│ ├── tools/
│ │ ├── index.ts # registerTools() — all 13 tools
│ │ ├── remember.ts # Store text + return extraction instructions
│ │ ├── add-person.ts # Create/deduplicate person
│ │ ├── add-org.ts # Create/deduplicate organization
│ │ ├── add-activity.ts # Log interaction with linking
│ │ ├── add-action.ts # Create follow-up/task
│ │ ├── link.ts # Create relationship between entities
│ │ ├── recall.ts # Search across all entity types
│ │ ├── people.ts # Query contacts
│ │ ├── orgs.ts # Query organizations
│ │ ├── activities.ts # Query interactions
│ │ ├── actions.ts # Query action items
│ │ ├── context.ts # Full briefing (person or org)
│ │ └── correct.ts # Update, delete, merge, reset
│ ├── extraction/
│ │ ├── prompt.ts # Regex extraction engine (legacy, used by seed script)
│ │ └── resolver.ts # Entity resolution (fuzzy match + Levenshtein)
│ ├── middleware/
│ │ └── auth.ts # API key auth (timing-safe comparison)
│ └── utils/
│ ├── config.ts # ~/.kernal/config.json management
│ └── format.ts # Response formatting helpers
├── bin/
│ └── kernal.ts # CLI: init, serve, status, export
├── dashboard/ # React dashboard (separate npm project)
│ ├── src/
│ │ ├── App.tsx # Shell with nav, command bar, view routing
│ │ ├── components/
│ │ │ ├── Overview.tsx # Stats grid + most connected + activity breakdown
│ │ │ ├── NetworkGraph.tsx # Force-directed graph (react-force-graph-2d)
│ │ │ ├── Timeline.tsx # Vertical activity timeline
│ │ │ └── ActionItems.tsx # Actions grouped by urgency
│ │ ├── hooks/
│ │ │ └── useKernalApi.ts # API fetch wrapper + types
│ │ └── lib/
│ │ └── intentRouter.ts # NL command → view routing
│ └── vite.config.ts # Proxies /api to cloud server
├── scripts/
│ └── seed-demo.ts # Seed demo data (12 people, 18 orgs, 19 activities)
├── test/
│ └── resolver.test.ts # 50 tests (resolution, extraction, tools, demo scenarios)
├── Dockerfile
├── package.json # @kernal/mcp
└── tsconfig.json
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.
- 2d ago First seen · 188 lines · 2,077 tokens per session scan A 49a6f7e19aab
kernal CLAUDE.md is an instructions file published in the GitHub repository pintomatic/kernal (2 stars, last pushed 2mo ago), licensed MIT. It adds 2,077 tokens to every session, about $0.0104 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.
Other instructions, from other repositories
m_flow AGENTS.md
AGENTS.md instructions for FlowElement-xinliuyuansu/m_flow, covering m-flow — developer & agent reference, 1. repository map, extension points, 2. local development and python backend (requires python 3.10 – 3.13).
engraphis AGENTS.md
Instructions for Coding-Dev-Tools/engraphis, covering agents.md — engraphis, 0. read this first — two architectures live in one package, 1. commands, ── unified dashboard + memory inspector ── and 2. the v2 recall pipeline (where the real work is).
engraphis CLAUDE.md
Instructions for Coding-Dev-Tools/engraphis, covering claude.md, the one rule that prevents most mistakes, before you say "done" — run the canonical gate, slash commands available here and working style in this repo.
Waggle-mcp AGENTS.md
Instructions for Abhigyan-Shekhar/Waggle-mcp, covering repository agent rules, custom rules and waggle automatic memory.
engram-mcp CLAUDE.md
Instructions for edg-l/engram-mcp, covering engram mcp, development rules, structure, key types and mcp capabilities.
Binder AGENTS.md
Instructions for mpazik/Binder, covering binder, tech stack, monorepo structure, testing and development.