Pedro Alvarado
Community@PedroAlvarado
Pedro Alvarado publishes jev-scout, a repository inspection skill that maps where TypeSafe Jev typed decision models replace repeated LLM judgments and brittle heuristics.
Agent Skills by Pedro Alvarado
Showing 1 vetted skills indexed across 1 GitHub repositories.
Frequently Asked Questions About Pedro Alvarado
FAQPage SchemaWhat tasks can I perform with Pedro Alvarado's jev-scout skill?▼
Jev-scout inspects a repository and identifies where TypeSafe Jev decision models add value: replacing repeated LLM judgments, first-match logic, fixed cutoffs, and defaults. It outputs a value map, decision inventory, ranked opportunities, decision contracts, and shadow-mode experiments.
Who is jev-scout designed for?▼
It targets engineers and product teams looking to reduce AI cost or latency, add typed routing, scoring, or verification, or explore what a system could do with frequent inexpensive decisions. It suits anyone asked to 'Jevify' a codebase or find Jev opportunities.
How does jev-scout run and what does it modify?▼
It runs on any Agent Skills host with shell and file access. The bundled scanner uses Bun or Node.js 22.6+ if installed, otherwise falling back to rg and git grep. It is analysis-only: it changes no code and writes nothing outside its report unless asked.
What license and cost apply to jev-scout?▼
Jev-scout is released under the MIT open-source license, so it is free to use, modify, and redistribute. It requires no Jev API key, and its only network reads are TypeSafe's public documentation, which are optional.
What prerequisites does jev-scout need?▼
You need an Agent Skills host with shell and file access. Bun or Node.js 22.6+ enables the bundled scanner; without either, the skill falls back to rg and git grep. No Jev API key or mandatory network access is required.