Skip to main content
Glama
jbt95

jev mcp

by jbt95
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Local

jev-toolkit

A small stdio MCP server for TypeSafe/Jev with three tools: typesafe_ask for structured judgments, typesafe_rank for ordering supplied candidates by relevance, and typesafe_verify for checking claims against supplied evidence. Calls use the TypeSafe System One API and are recorded in a local event log.

Built with Effect and TypeSafe's SDK. Bun >= 1.3.14 is required.

Install

git clone git@github.com:jbt95/jev-toolkit.git
cd jev-toolkit
bun install --frozen-lockfile
scripts/install.sh
export TYPESAFE_API_KEY=…

Register Jev in a user-level harness config with:

jev install opencode
jev install pi
jev install omp
jev install claude-code

The Pi command installs pi-mcp-adapter and requires the Pi CLI and network access. All commands preserve unrelated MCP settings and set JEV_HARNESS automatically. They do not store TYPESAFE_API_KEY; make sure the harness process provides it to the MCP server. See docs/mcp.md for configuration paths and tool schemas.

Related MCP server: jevcore-mcp

Development

The four gates are:

bun run lint
bun run format:check
bun run typecheck
bun run test

The application and CLI run on Bun; the Oxlint RuleTester fixture suite uses Node >= 22.18 because that external test API is Node-only. Tests are offline: they use fake transports, temporary files, and localhost only. TYPESAFE_API_KEY is read at call time and never logged. Verification redacts credentials and never writes claim or evidence text to the event log; rank rejects common code, diff, and transcript patterns, redacts credentials, and never logs query or candidate text. Its pattern checks are conservative, not a complete content classifier. Ask state is caller-provided and should be redacted by the caller.

This server cannot be deployed

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Maintenance

ActivityMaintained
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers

  • Jev

    mywwave
    A
    license
    Not graded
    quality
    C
    maintenance
    Enables Cursor agents to route subagent tasks and gate reads, writes, and shell commands through fast TypeSafe Jev judgments, exposing tools for choice, scoring, and noul decisions.
    1
    MIT

  • jevcore-mcp

    PerryLink
    A
    license
    Not graded
    quality
    B
    maintenance
    Enables MCP hosts to query Jev's typed decision model—yes/no, choice, and score—with calibrated probabilities, while defaulting to an offline mock and disclosing all egress unless explicitly enabled.
    31
    Apache 2.0

  • Jev MCP

    resumocast
    A
    license
    Not graded
    quality
    A
    maintenance
    Enables MCP clients to submit bounded semantic-uncertainty judgments to the pinned TypeSafe Jev API, with tools for yes/no, choice, and score evaluations plus optional evidence or context selection.
    MIT

  • jev_mcp

    wr0x00
    A
    license
    Not graded
    quality
    B
    maintenance
    Provides MCP tools that wrap Jev's finite typed judgments to give agents a semantic decision-control layer for task routing, pre-execution risk gating, and completion verification, keeping final execution authority in code.
    MIT