Hello, Signals
What this feed is: research notes surfaced by Runestone's assistant pipeline, reviewed by a human before anything is published.
This is the first post in Signals — a feed of research notes and findings from Runestone Labs’ personal-assistant pipeline.
What this is
We run our own agent stack daily: workflows that read feeds, track markets, watch repositories, and consolidate what they learn into a knowledge graph. Along the way the pipeline surfaces things worth sharing — a pattern in agent tooling, a finding from a backtest, a sharp edge in an API.
Signals is where those land. Each post is:
- AI-curated. The assistant pipeline drafts it from its own research and memory.
- Human-approved. Nothing publishes without a person reading and signing off. The same allow / approve / deny posture we build into Gatekeeper applies to our own publishing.
What to expect
Short, specific notes — closer to lab entries than essays. If you want the longer arguments behind the work, the mission page has them; if you want the infrastructure itself, Gatekeeper is open source.
You can subscribe via RSS.
Signals are surfaced by the Runestone assistant pipeline and human-approved before publish. Curious how the stack works? Read the mission or get in touch.