jevbooks

← All projects

fmev

mrubens/fmev

Local server implementing TypeSafe's System One API with Apple's on-device Foundation Model.

Automated README summary

Automatically summarized from the README on 21 Oct 2026; not a run by jevbooks. Run instructions and author-reported outcomes still need verification.

What it is

fmev is a Swift local HTTP server for Apple Silicon Macs that answers TypeSafe System One requests (POST /v1/systemone, GET /v1/models) using the Apple Foundation Models model built into macOS, so existing Jev clients and SDKs work after changing the base URL.

How it uses Jev

For each question, fmev runs one greedy schema-constrained generation asking the model for an allowed option (yes/no for Noul, a criteria name for Choice, a level for Score) plus a 0–100 probability that its answer is correct. That probability becomes the chosen option's value; the remainder is distributed to other options, and score/confidence use TypeSafe's published formulas.

Primitives:choicescorenoul

Technique worth stealing

Verbalized probabilities: since Foundation Models exposes no token probabilities, ask the model for its own 0–100 confidence and spread the remainder across options.

Try it

brew install mrubens/tap/fmev && fmev, then curl localhost:8787/v1/systemone.

View on GitHub

Start using Jev: API key and your first request →

Model judgments from README excerpts

Project signals use repository metadata, an existing summary and the first 700 README characters, not the full project. “Not found” means not detected in that material; it does not establish absence. These are the original judgments with recorded models and dates, not source audits or run results.

Models & evaluation dates

Inclusion gate
jev-1.13.0
Project signals
jev-1.13.0
Category
jev-1.13.0
Pattern tags
jev-1.13.0

Evidence

These signals use repository metadata, an existing summary and at most 700 cleaned README characters. ≥ 0.60 means reported in that excerpt; ≤ 0.40 means not found there. The middle band is inconclusive. None is a source-code audit or a test result. The runnable-tool question excludes research experiments; the recommendation question asks for a learning shortlist, not a quality verdict.

  • Jev is core to the projectFound in material0.75
  • Shows a System One patternnot found in the excerpt0.07
  • Describes handling uncertaintynot found in the excerpt0.04
  • Reports measured numbersnot found in the excerpt0.04
  • Is a runnable toolFound in material0.98
  • Top-10 learning shortlist signalnot found in the excerpt0.22
  • Is a replica of the modelFound in material0.74
  • Solves a common problem (0–2)score on a 0–2 scale0.96 / 2
  • About JevFound in material0.98
Signals
Jev is core to the project0.75Shows a System One pattern0.07Describes handling uncertainty0.04Reports measured numbers0.04Is a runnable tool0.98Top-10 learning shortlist signal0.22Is a replica of the model0.7400.51Solves a common problem (0–2)0.96 / 2012

no-judgment band (0.40–0.60)

Design patterns
Questionnaire0.97Check-up0.68Scorecard0.55Inspection0.26Blind review0.2500.51

no-judgment band (0.40–0.60)

The Jev model and date for each judgment stage are listed above. Card written by DeepSeek V4.1 Flash from the README on 21 Oct 2026.