Two folders. One tells the agent what stack to reach for before cargo new. The other measures the repo against a fixed gate registry and writes the plan down. Trust, Security, and Incentive — the agent gets the same contract the house uses.
The playbook for starting a Rust app, service, or website that can ship onto the MATA distributed cloud the day it goes live.
Answers "what do we use for X" with a fixed stack: mID for identity, SpaceDB for storage, remade_ffmpeg_rs for media, FFAI for AI, rusty_alloc for allocation, Deputy for the supply chain. Architecture, deploy-readiness seams, performance doctrine, and the build/validate workflow live in the same folder. Read it before the first dependency, not after the rewrite.
Open the skill folder→
A security-auditing framework for safe code — 41 gates, a plan file per unit, a README table that does not lie.
The agent audits every crate against a fixed registry, writes docs/plans/use-protection-please.md in identical structure, and regenerates Incomplete / Scheduled / Completed from that file. Evidence, not vibes. Rust-first — cargo-deny, cargo-audit, Miri, fuzz, Kani — and language-agnostic enough to run on any repo you point it at.
Open the skill folder→
