Vertical Agent Packs

Plug-in AI agents
that run on your hardware

Standalone, configurable agent packs built from the Daily AI OS. Install on any Mac or Linux box with Ollama or LM Studio. No cloud APIs required.

Local-first Ollama + LM Studio Human-in-the-loop One-time purchase MIT license

No cloud lock-in

Runs on your hardware via Ollama or LM Studio

Approval-gated

Nothing publishes or executes without your sign-off

Standalone

Each pack works independently — no OS dependency

MIT licensed

Use commercially, modify freely, no royalties

Three packs. Three workflows.

Extract the highest-leverage parts of the AI OS into standalone tools you can ship on any local setup.

Content Pipeline
✍

Content Pipeline Pack

Draft, fact-check, and publish content using local AI. Blog posts, X threads, newsletters — generated and quality-gated before you approve.

  • 4 content formats: blog, thread, newsletter, product
  • Quality gate catches hallucinated metrics and round-number claims
  • Optional Twitter/X publishing via OAuth 1.0a
  • Customizable templates and persona/tone config
  • Draft storage with YAML frontmatter and publish tracking
content_generator.py quality_gate.py publisher.py 4 templates 3 examples
$99 one-time or $29/mo
Get Pack
Trading Signals
📈

Trading Signals Pack

Daily AI signal reports and weekly performance summaries from paper trading strategies. Runs fully locally — paper trading only by default.

  • 3 strategy engines: momentum, mean reversion, trend-following
  • Risk metrics: Sharpe, Sortino, max drawdown, win rate
  • AI narrative commentary on each report
  • yfinance integration or local CSV price data
  • Paper trading portfolio tracker with P&L history
signal_reporter.py 3 strategies 2 templates 2 examples
$99 one-time or $29/mo
Get Pack
Research Agent
🔍

Research Agent Pack

Three agents that turn questions into structured research reports and build a searchable local knowledge base over time.

  • Quick researcher: question → structured report in 3-5 min
  • Deep researcher: multi-pass with local file context
  • Knowledge indexer: searchable local knowledge base
  • 4 templates: research report, comp analysis, market research, decision memo
  • Codebase analysis: feed source files as research context
researcher.py deep_researcher.py knowledge_indexer.py 4 templates
$99 one-time or $29/mo
Get Pack
Setup

Running in four steps

1

Install a backend

Start Ollama or LM Studio. Pull a model (qwen2.5:14b recommended). No GPU required — runs on Apple Silicon.

2

Run install.sh

The installer checks your backend, creates directories, copies the config template, and optionally pulls the recommended model.

3

Edit config.json

Set your model preferences, quality thresholds, output directories, and optional credentials. Every setting has a documented default.

4

Run an agent

python3 -m agents.researcher "your question" — results appear in the terminal and are saved to your output directory.

Comparison

Pack vs. full OS

Capability Pack Full OS
Core agent functionality Full Full
Standalone, no OS dependency Yes No
Telegram command interface Not included Included
Exocortex / memory system Not included Included
Multi-agent pipeline orchestration Not included Included
Approval queue / governance layer Terminal only Full Telegram flow
Site auto-publishing File export only GitHub Pages injection
Metrics and weekly reports Not included Included
Setup complexity Low (install.sh) Medium
Get Started

One purchase. Runs forever on your hardware.

No subscriptions required. No API keys expiring. No vendor raising prices. Buy once, run on any machine with Ollama or LM Studio.

Get All Three Packs — $249 View Full OS

Individual packs $99 each · $29/mo subscription available · MIT license