kernel CLAUDE.md

kernel CLAUDE.md is an instructions file for coding agents from isaacsight/kernel. It costs 812 tokens per session, scanned A, original, MIT.

Claude Code instructions for isaacsight/kernel, covering claude.md — kernel.chat group, what you need to know in 30 seconds, where to look for what and five rules that always apply.

Instructions file

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/isaacsight/kernel/claude-md
Clone the repo
git clone --depth 1 https://github.com/isaacsight/kernel

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.

agentmods badge for kernel CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/isaacsight/kernel/claude-md.svg)](https://agentmods.dev/instructions/isaacsight/kernel/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/isaacsight/kernel/claude-md"><img src="https://agentmods.dev/badge/instructions/isaacsight/kernel/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 812 This file is loaded in full into every session.
When invoked 812 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin unknown 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.

ModelPer sessionOnce invoked
Fable 5 $0.00812 $0.00812
Opus 5 $0.00406 $0.00406
Sonnet 5 $0.00162 $0.00162
Haiku 4.5 $0.00081 $0.00081

Measured today against content hash 7ffe8a659b84, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

kernel 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 today.

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 · 66 lines

How it starts

The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md — kernel.chat group

This file is a shim. The canonical project reference is KERNEL.md. Read that first.

CLAUDE.md exists because Claude Code auto-loads it at session start; the canonical reference moved to KERNEL.md on 2026-04-29 when the project's two surfaces (kbot agent + kernel.chat magazine) became co-equal and the older K:BOT-as-product + web-as-companion framing stopped describing the work.

What you need to know in 30 seconds

Two surfaces, one publication:

  1. kbot — open-source terminal AI agent (@kernel.chat/kbot on npm). Currently v4.5.0. ~100 specialty skills curated from 670 in v4.0 with a public audit trail. Local-first, BYOK, MIT.
  2. kernel.chat — editorial magazine. Currently on ISSUE 426 — THE WEATHER OVER THE QUEUE. Sixty-seven issues in the catalog. POPEYE-grammar anchored, PAPERSKY + WIRED decoded as editorial neighbours.

Discipline (same on both surfaces): count what gets read; cut what doesn't; file the audit in public; keep the manuscripts in the drawer. The room is different, the job is the same.

Where to look for what

If you need… Read
Project shape, current state, directory map, ship flow KERNEL.md
Magazine publishing workflow src/content/issues/PUBLISHING.md
Magazine visual grammar docs/design-language.md
Unified creative canvas handoff and agent-control contracts docs/creative-canvas-handoff.md
kbot v5 futures substrate packages/kbot/V5_FUTURES_PLAN.md
Why the 4.0 cut went the way it did packages/kbot/RELEASE_NOTES_4_0.md
Per-session working memory SCRATCHPAD.md

Five rules that always apply

  1. Magazine vocabulary in user-visible copy. Issue / feature / spread / folio / monument / colophon / dateline / postmark — never dashboard / panel / card / widget / modal. Never name "POPEYE" on the site.
  2. Evidence-cited commits. Every release commit cites numbers and reasoning; audit trails (CSV, JSONL, RELEASE_NOTES) ship alongside code changes.
  3. BYOK is the contract. kbot never hardcodes a provider preference. Local-first when there's a free path (Ollama/LLaDA/MLX) for the task.
  4. No emojis in code or user-visible copy unless the user explicitly asks (single-glyph system asterisk ★ is the exception, ratified in ISSUE 370).
  5. Update SCRATCHPAD.md at session end. Future sessions inherit your context through it.

Read the full file on GitHub · 66 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.

  1. today First seen · 66 lines · 812 tokens per session scan A 7ffe8a659b84

Subscribe to this mod's changes

kernel CLAUDE.md is an instructions file published in the GitHub repository isaacsight/kernel (16 stars, last pushed 5d ago), licensed MIT. It adds 812 tokens to every session, about $0.0041 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 2026-09-03.