Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

jev playground

Weekly go-to-market workflows built on Jev, such as lead triage and deal risk, each tested against Claude.

Benchmarkprototype
View on GitHub ↗

Overview

jev playground by deepdave98 is a series of practical go-to-market workflows built on TypeSafe's Jev and tested against Claude, published roughly weekly. Topics so far are lead triage and routing, deal risk, reply triage, MEDDPICC extraction from call summaries, duplicate account detection, closed-lost deal reactivation, and an MCP server.

The list summaries say each workflow is benchmarked with its harness published so the comparison can be checked. The top-level README is just an index, so the details of questions, datasets and results live in each week's folder. No licence file is included.

What the decision model decides

TypeQuestionAnswers
ChoiceWhere should this lead be routed?(inferred)—
ScoreHow risky is this deal?(inferred)—
Yes / noAre these two accounts duplicates?(inferred)yes · no

How it works

  1. Pick a GTM task such as lead triage
  2. Frame it as typed Jev questions
  3. Run the same task with Claude
  4. Compare results in the published harness

Strengths

  • Concrete sales-ops use cases
  • Side-by-side comparison with an LLM
  • Harnesses published per workflow

Caveats

  • No licence file
  • Top-level README gives no results; details are per folder

At a glance

Type
Benchmark
Status
prototype
Runtime
Scripts per workflow plus an MCP server
Has code
yes

Use cases and integrations

How curated lists describe it

  • Weekly builds on Jev (TypeSafe AI), benchmarked honestly enough to publish.

    hellogumbo/awesome-jev

  • Weekly go-to-market workflows built on Jev, such as lead triage and deal risk, each tested against Claude with its harness published. No licence file.

    MrJev/awesome-jev

Frequently asked questions

What workflows are covered?

Lead triage, deal risk, reply triage, MEDDPICC, account dedup and deal reactivation, plus an MCP server.

Is Jev compared with anything?

Yes. Each workflow is tested against Claude.

Can I reuse the code?

There is no licence file, so reuse terms are unclear.

Related projects

  • jev-experiments

    Experiments testing Jev ideas against baselines, starting with Jev as a top-10 tool router for MCP agents on three benchmarks.

    ProjectresearchPython
  • playwright-cli-jev-bakeoff

    Reproducible bake-off of Playwright CLI alone, CLI plus Jev, and Playwright MCP on fuzzy browser goals, with honest results.

    MCP serverresearchPython
  • jev-mcp (arunav25)

    MCP server exposing Jev as one evaluate tool, plus a harness comparing Jev's judgments with general LLMs on shared datasets.

    MCP serverbetaJavaScript
  • typesafe-jev-tools

    Claude Code hook that, when you write an LLM call or classifier, injects a three-way test: plain code, frontier model or Jev.

    HookbetaBash
  • jev-browser (MahmoudAdelbghany)

    Jev-powered browser MCP server that makes each step in one parallel call and escalates to the driving LLM only when unsure.

    BenchmarkbetaTypeScript
  • toolJev

    Code Mode for MCP where retrieval shortlists tools and Jev picks one or says none fits, benchmarked on several tool-use suites.

    BenchmarkresearchPython