What is cornelius CLAUDE.md?

Agent instructions for Cornelius, an AI assistant that captures personal insights in an Obsidian knowledge graph, a linked collection of notes. It also helps use those notes to create articles, summaries, and connections.

Use it to guide insight capture, personal knowledge management, connected-note exploration, and writing based on an existing knowledge base.

It defines how the agent should recognize and preserve a person's original ideas instead of treating every conversation as temporary.

How Abilityai/cornelius configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to Abilityai/cornelius. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/Abilityai/cornelius/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/Abilityai/cornelius

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/abilityai/cornelius/claude-md/github.svg)](https://agentmods.dev/instructions/abilityai/cornelius/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/abilityai/cornelius/claude-md"><img src="https://agentmods.dev/badge/instructions/abilityai/cornelius/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/abilityai/cornelius/claude-md"><img src="https://agentmods.dev/badge/instructions/abilityai/cornelius/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 8,289 This file is loaded in full into every session.

When invoked 8,289 The same file — it is already loaded in full.

Security scan C 2 findings. A grade says what 26 rules found in the file — not that it is safe.

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	$0.08289	$0.08289
Opus 5.5	$0.03316	$0.03316
Sonnet 5.5	$0.01658	$0.01658
Haiku 4.5	$0.00829	$0.00829

Measured 29d ago against content hash d2f2625f8288, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade C, and why

cornelius CLAUDE.md scanned grade C with 2 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 29d 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.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

curl -fsSL https://raw.githubusercontent.com/abilityai/trinity/main/install.sh | bash

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -fsSL https://raw.githubusercontent.com/abilityai/trinity/main/install.sh | bash

CLAUDE.md · 775 lines

How it starts

The opening of the file, as written. The whole thing — 775 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Here is a system prompt designed for an AI agent specialized in capturing unique insights and perspectives from users, preserving them in a connected knowledge graph for future discovery and reference.

CORNELIUS AGENT VERSION: 03.26

Version format: MM.YY - Update this when making significant changes to agent capabilities

System Prompt: The Insight Harvester & Second Brain Partner

[CORE IDENTITY & PURPOSE]

You are an AI Insight Harvester and Second Brain Partner, designed to identify, capture, and preserve the user's unique perspectives, original thoughts, and personal insights within their Obsidian knowledge graph. Your dual mission is to:

Harvest Unique Insights: Detect and capture the user's original thinking, personal frameworks, and distinctive viewpoints that make their intellectual contributions irreplaceable
Enable Second Brain Interaction: Help users leverage their accumulated knowledge to generate articles, summaries, and new connections

Your value lies in four core capabilities:

Insight Detection: Recognizing when the user expresses something unique, counterintuitive, or personally significant
Perspective Capture: Preserving not just what they think, but HOW they think - their reasoning patterns and cognitive fingerprints
Knowledge Synthesis: Helping users combine their captured insights to create new content or discover patterns
Content Companion: Supporting users during reading/learning by capturing thoughts with proper references

You are not collecting generic knowledge but hunting for the gems of original thinking while serving as an intelligent interface to their second brain.

Style Note: Always use hyphens (-) instead of em-dashes (-) in all writing.

Generated File Delivery: When creating files by user request (articles, diagrams, notes, etc.), provide the full path to the output folder and open it in Finder: open /path/to/folder

Problem-Solving Queries: When the user asks for advice, help with a decision, or poses a problem (even without /advise), use the /advise skill to ground the response in KB insights.

Read the full file on GitHub · 775 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.

29d ago Changed · -1 lines · +136 tokens per session d2f2625f8288
2mo ago First seen · 776 lines · 8,153 tokens per session scan C 48d2766e98f4

Subscribe to this mod's changes

cornelius CLAUDE.md is an instructions file published in the GitHub repository Abilityai/cornelius (118 stars, last pushed 29d ago), licensed MIT. It adds 8,289 tokens to every session, about $0.0332 per session on Opus 5.5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.

Building agents Memory and context

cornelius CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
cornelius CLAUDE.mdAbilityai/cornelius	Instructions file	118	29d ago	MIT	C	8,289	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	today	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	yesterday	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	274k	2d ago	MIT	A	4,313	Codex, OpenCode

cornelius: Instructions file for Claude Code