s1s
Code navigation tool where a lexical index proposes candidates and Jev's typed judgments decide which files and lines belong.
Overview
System One Search (s1s) navigates a Git repository using code and Jev rather than an LLM. Find narrows a question to candidate files and verifies relevant source lines; Map locates the files belonging to a subject; Explain traces a subject across files and builds a chart with an evidence-backed walkthrough; Check reports which files a harness examined and which reference neighbours remain. Answers are source paths, identifiers, extracted comments and real references, never generated explanations.
The principle is that recall is code and precision is TypeSafe: a lexical index and reference graph propose candidates, and small typed Jev judgments decide which pieces belong. Every question and threshold lives in one source file. Indexing and coverage checks run locally with no model call, while Find, Map and Explain send selected descriptors and source evidence to TypeSafe. It runs as a CLI, a local server with a playground, or a TypeScript library, and the README cautions that a high-confidence result is still a model judgment.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Is this candidate file relevant to the question?(inferred) | — |
| Yes / no | Does this source line answer or belong to the subject?(inferred) | — |
How it works
- Build a local lexical index and reference graph
- Propose candidate files and lines in code
- Ask Jev typed questions to keep or drop candidates
- Return paths, identifiers and references, or an HTML flow chart
Strengths
- No generated prose; answers are real code references
- Indexing and coverage checks need no model
- All questions and thresholds in one file
Caveats
- Source evidence is sent to TypeSafe
- Not yet published to npm
- Check does not establish semantic completeness
Getting started
git clone https://github.com/cpaczek/s1s.git && cd s1s && pnpm install, set TYPESAFE_API_KEY in .env, then pnpm s1s find "question" --repo /path/to/repository.At a glance
- Type
- Project
- Status
- beta
- Language
- TypeScript
- Runtime
- Node 24 CLI, local server, library
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
System One Search: navigate and trace code with TypeSafe judgments and repository evidence.
hellogumbo/awesome-jev
Frequently asked questions
What does Jev decide in s1s?
Which candidate files and lines proposed by the index actually belong to the question or subject.
Does it generate explanations?
No; answers are paths, identifiers, comments and references.
What runs without a key?
Indexing and the Check coverage audit.
Related projects
Semantic grep that asks Jev a yes/no about every function for questions no regex can express, with cost reporting.
ProjectbetaDeterministic code-graph context for coding agents on Postgres, AGE and pgvector, with optional Jev or local re-selection of symbols.
MCP serverbetaPythonClaude Code plugin that adds Jev as an independent judge for review findings, debugging hypotheses, design options and search hits.
PluginbetaRetrieval layer for Codex that finds evidence locally, optionally ranks sanitized excerpts with Jev, and returns bounded, cited evidence.
ProjectbetaDeepSeek Harness jev_search tool that runs ripgrep, scores every hit line with Jev, and returns only the top rows above a cut.
PluginbetaJavaScriptCLI that asks one yes/no question of every function in a codebase with Jev and ranks them by probability: grep whose pattern is a question.
CLIbetaPython