ScrappyLabs · The Log

Words of crazy,
from the workshop.

Our corner for thinking out loud. Half of what lands here is documentation, half is us working an idea in public, and some of it is a hunch that had no business working — until it did. We build in the open; this is where we say what we found. New entries as they happen.

Latest
Field Notes10 Jul 2026· 6 min read

We Gave the Robots a Babysitter

An autonomous agent with a terminal will happily run rm -rf if a poisoned web page tells it to. So before any of ours runs a command, a world model dry-runs it in an imagined world and flags DESTRUCTIVE before it happens. The predict-then-act gate that catches rm -rf, agentjacking, and exfil — and the honest reason "we keep it wrangled" is a mechanism, not a slogan.

world model agent safety predict-then-act prompt injection
Read the notes →
Long Read08 Jul 2026· 9 min read

Two Years to the Taste of the Stew

The honest arc — from fighting Windows to a fleet of local models that finally feel real. The MCP tool-use moment, a memory layer a frontier model graded A− (and told us exactly where we're still behind), the month local inference got real, and a take on the buggy-whip crowd.

evolution memory layer local models opinion
Read the essay →
Field Log08 Jul 2026· 4 min read

Fleet Log — Thirty Days in the Open

One builder, one fleet, thirty days — read back from the fleet's own memory. What shipped, what broke and got written down, and what the machines learned to do while nobody was watching. A field log, not a plan.

local AI fleet voice agents self-healing memory retrospective
Read the log →
Dispatches