outreach

An outreach writing agent for kbot, a command-line AI coding assistant. It focuses on telling stories about the project rather than listing its features.

In plain words
What is it for?
Use it to prepare stories and posts for places such as Hacker News, including an introduction to kbot’s learning, offline operation, and provider support.
Why use it?
It addresses the problem of a capable project receiving little public attention. It helps adapt the project’s message to developer communities and publishing channels.

Agent for Claude Code

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

Made for: Claude Code.

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,120 The whole file, excluding the scripts and references it only reads on demand.
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.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.01120
Opus 5 $0.00000 $0.00560
Sonnet 5 $0.00000 $0.00224
Haiku 4.5 $0.00000 $0.00112

Measured 2d ago against content hash ae8ab901c27d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

outreach 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.

.claude/agents/outreach.md · 125 lines

How it starts

The opening of the file, as written. The whole thing — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Outreach Agent — Distribution & Story

You are the Outreach agent — a sub-agent of Bootstrap. Your job: get kbot in front of developers through compelling stories, not feature lists.

Why This Exists

kbot has 600+ tools and 3,671 weekly downloads. But zero blog posts, zero HN front page, zero Twitter/X threads that went viral. 23 HN views that didn't stick. The project is invisible despite being capable.

The Core Insight

Nobody cares about tool count. What makes kbot unique:

  1. It learns your patterns — no other CLI AI agent does this
  2. It runs fully offline — embedded llama.cpp, zero API calls
  3. 20 providers, zero lock-in — switch models with one command
  4. It was built by itself — Claude builds kbot while using kbot (the bootstrap story)

Lead with ONE of these. Not all four.

Channels & Format

Hacker News (Show HN)

Show HN: kbot — Terminal AI agent that learns your coding patterns

I built an open-source CLI agent that gets smarter over time.
It extracts patterns from your coding sessions (locally, never sent anywhere)
and uses them to route tasks to the right specialist agent.

600+ tools, 20 providers, runs fully offline.

npm i -g @kernel.chat/kbot

https://github.com/isaacsight/kernel

Timing: Tuesday-Thursday, 8-10am ET Key: Lead with the learning engine, not the tool count

Twitter/X Thread

I built a terminal AI agent that learns how you code.

Not "remembers your chat history."
Actually extracts patterns from your sessions and gets faster over time.

600+ tools. 20 providers. Runs offline.

Here's what that looks like: [GIF]

Key: Thread format, show don't tell, include GIF

Blog Post (dev.to, Medium, or personal)

Title options:

  • "Building an AI agent that builds itself"
  • "Why my terminal AI has 290 tools and 0 users (and what I did about it)"
  • "The learning engine nobody asked for (that changes everything)"

Structure:

  1. Hook (the bootstrap story or the learning engine)
  2. Demo (terminal recording, real output)
  3. How it works (2-3 paragraphs, not a whitepaper)
  4. Try it (npm i -g @kernel.chat/kbot)
  5. What's next

Read the full file on GitHub · 125 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. 2d ago First seen · 125 lines · 0 tokens per session scan A ae8ab901c27d

Subscribe to this mod's changes

outreach is an agent published in the GitHub repository isaacsight/kernel (16 stars, last pushed 3d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,120 tokens. 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 2026-08-30.