typesafe-mod
Claude Code function hook that sends the agent's this-or-that questions to Jev and optionally ranks installed skills per prompt.
Overview
typesafe-mod is a Claude Code function hook, a TypeScript module running inside the agent loop, that routes two kinds of decision to Jev. By default, when the agent stops to ask you a this-or-that question through AskUserQuestion, the hook sends it to Jev and writes each option's probability into the transcript. You still choose, unless TYPESAFE_AUTO_ANSWER=1 lets Jev answer when confident. Multiple questions ride one request and are judged independently, so Jev can be confident about one and hand the next back.
An optional skill router on prompt submit ranks every installed skill against your prompt and asks whether the turn needs a procedure at all; a confident winner becomes one advisory line of context, preserving prefix caching. It is off by default for cost reasons. It requires early-access function hooks to be enabled and a TypeSafe key.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Choice | Which option would the user pick for the agent's question? | — |
| Choice | Which installed skill fits this prompt? | — |
| Yes / no | Does this turn need a procedure (skill) at all? | yes · no |
How it works
- The agent calls AskUserQuestion or a prompt is submitted.
- The hook sends options or skills to Jev in one request.
- Probabilities are shown; above a confidence floor Jev can answer.
- Below the floor the question goes back to the user.
Strengths
- Answers routine agent questions automatically when enabled.
- Per-question confidence floor keeps uncertain choices with the user.
- Skill hint preserves prefix caching.
Caveats
- Requires early-access Claude Code function hooks.
- Skill routing adds per-prompt cost and is off by default.
Getting started
claude plugin marketplace add BeLazy167/typesafe-mod && claude plugin install typesafe-mod@typesafe-modAt a glance
- Type
- Project
- Status
- beta
- Language
- TypeScript
- Runtime
- Claude Code function hook/plugin
- Has code
- yes
Use cases and integrations
How curated lists describe it
Claude Code mod that routes decisions to TypeSafe's Jev model: ranks installed skills per prompt, and answers the agent's own this-or-that questions when confident.
hellogumbo/awesome-jev
Frequently asked questions
Will it answer questions for me?
Only with TYPESAFE_AUTO_ANSWER=1 and when Jev's confidence clears the floor.
What is the skill router?
An optional ranking of installed skills per prompt, off by default because of cost.
What is required?
Claude Code function hooks enabled and a TypeSafe API key.
Related projects
OmO/senpi agent plugin where Jev or Span-01 rates which skills and next tools fit the task and spots loops or completion.
Agent skillbetaTypeScript- SkillRanker★ 134
Rust CLI that ranks which agent skills fit the next step from live session context with Jev, can abstain, and hooks into Claude Code.
CLIbetaRust DeepSeek Harness plugin that registers a native jev_decide tool and declares its own policy for when the agent should ask Jev.
PluginbetaJavaScriptOne skill and MCP server for Claude Code, Codex and Hermes where Jev gates large-model turns, routes models and guards Git steps.
MCP serverbetaPythonOpen-source Jev skills for Claude Code, Codex, Cursor and Grok: one client to gate, route, guard, verify, triage, review and act.
ProjectbetaTypeScriptLocal gateway and MCP tools that let Jev predict which tool a coding agent should call next, from suggestions to enforced choices.
MCP serverbetaTypeScript