jevsume
Resume review app and MCP server where a Cloudflare Worker asks Jev typed questions and composes a JevScore, generally or per job.
Overview
jevsume reviews resumes for ATS-friendliness using Jev. The frontend extracts resume text the way an applicant tracking system parser would, then a Cloudflare Worker runs typed Jev questions and composes a JevScore. General review covers wording, conciseness, structure, metrics and parseability. Per-job review turns a job description and tags into a stored job persona, the state and questions for that job, and scores one resume against it.
Jev produces scores, verdicts and probabilities; the Worker turns them into UI copy. The same review is exposed as an MCP server, hosted without auth but rate-limited, or run locally with your own key via npx, with tools to list, fetch and suggest job lenses and to review a resume. An agent skill describes a loop of picking a lens, reviewing, rewriting weak bullets and re-reviewing, and warns agents not to invent Jev scores. The stack is a React SPA and Hono API on Cloudflare Workers with D1 storage.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Score | How well does this resume perform on wording, conciseness, structure, metrics and ATS parseability? | — |
| Score | How well does this resume match this job persona? | — |
| Choice | Which job lens best fits this resume? (suggest_job_lens) | — |
How it works
- Extract resume text as an ATS parser would
- Choose general review or a job persona
- Worker sends typed questions to Jev
- Scores are composed into a JevScore with findings and gaps
- Agents can loop via MCP: review, rewrite, re-review
Strengths
- Mirrors ATS text extraction
- MCP tools return compact JSON
- Hosted endpoint needs no key; local mode uses yours
Caveats
- Hosted MCP has no auth yet and a 10 per minute limit
- Scores are model judgments, not hiring outcomes
- Resume text is sent to the API
Getting started
npx -y github:unownone/jevsume -- --api-key $TYPESAFE_API_KEY (local MCP), or pnpm dev for the app.At a glance
- Type
- App
- Status
- beta
- Language
- TypeScript
- Runtime
- Cloudflare Worker, React SPA, MCP server
- Has code
- yes
Use cases and integrations
How curated lists describe it
A resume-review app that checks general writing and structure or compares a resume with a specific job description.
logicrw/awesome-jev-projectsATS-friendly resume review powered by Jev (TypeSafe System One). The frontend extracts resume text the way a parser would, then a Cloudflare Worker runs typed JEV questions and composes a JevScore.
hellogumbo/awesome-jev
Frequently asked questions
What does Jev score?
Resume quality dimensions and fit against a job persona, combined into a JevScore.
Can my agent use it?
Yes, via the hosted or local MCP server and the bundled skill.
Does it write my resume?
No. Jev only scores; an agent using the skill can rewrite bullets and re-review.
Related projects
Local long-term memory for coding agents with an optional Jev reranker that orders recalled memories.
ProjectproductionTypeScriptAgent skill where Jev flags 35 AI-writing patterns per paragraph and the model edits only the flagged spots.
ProjectbetaPythonAgent skill where Jev scores a task prompt's clarity, the agent refines it up to three times until it reaches 80/100, then builds.
ProjectbetaJavaScriptCLI that scores every test against a git diff with Jev and emits filter arguments for vitest, jest, node, Playwright, cargo, go and rspec.
ProjectbetaTypeScriptMCP server that returns only the JS/TS code a task needs by following the TypeScript compiler graph and asking Jev about relevance.
MCP serverbetaTypeScriptLocal-first MCP server that returns only Jev-checked passages from big PDFs, Word files and spreadsheets, with exact row calculations.
MCP serverbetaTypeScript