发现与本库做同一件事的 agent skill：`decision-first`
Agent skill that makes a coding agent try Jev first for bounded judgments like classify, rank or gate, and log each attempt.
Overview
decision-first is an agent skill. Once installed, the coding agent notices when a step is really a bounded judgment (classify, tag, triage, size, rank, route, filter, verify, match, gate) and tries TypeSafe's Jev before writing a regex, a keyword heuristic, an LLM prompt or doing it by hand.
It then documents the attempt, adopted or not, in a decision-lab directory so later projects can reuse what was learned. It works with Claude Code, Codex, Cursor, Gemini CLI, OpenCode, Pi, Hermes or any agent that can run shell commands and load skills, and is careful about keys: the agent must never ask you to paste one into chat or write it to a file.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Other | Is this step a bounded judgment that Jev could answer as a Choice, Score or Noul?(inferred) | — |
How it works
- Agent spots a bounded judgment in the task
- It drafts typed questions and calls Jev via the skill's scripts
- It compares the result with the alternative approach
- It records the case, adopted or not, in the decision lab
Strengths
- Encourages cheap typed decisions over brittle heuristics
- Builds a reusable log of what worked
- Safe key handling rules
Caveats
- Requires python3 and a TypeSafe key
- Behavioural skill; results depend on the host agent
Getting started
npx skills add harrymunro/decision-first --skill decision-first (or claude plugin marketplace add harrymunro/decision-first && claude plugin install decision-first@decision-first)At a glance
- Type
- Agent skill
- Status
- beta
- Language
- Python
- Runtime
- agent skill / Claude Code plugin
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
Frequently asked questions
What does the skill change?
It makes the agent try Jev first whenever a step is a bounded judgment, then log the outcome.
Where are results logged?
In a decision-lab directory, ~/Workspace/decision-lab by default or $DECISION_LAB.
Which agents are supported?
Claude Code, Codex, Cursor, Gemini CLI, OpenCode, Pi, Hermes and others that load skills.
Related projects
- Building with Jev★ 155
Agent skill for writing and improving programs that call Jev: question design, state structure, thresholds and diagnosing wrong answers.
Agent skillproduction MCP server giving coding agents fast typed decisions (ship gates, risk checks, file triage, page checks) as accept, review or escalate.
MCP serverbetaTypeScriptThirty worked Jev uses in ten levels, from a smart if statement to a pi coding agent that writes its own Jev questions.
ProjectbetaTypeScriptCLI and MCP server giving coding agents Jev-powered semantic search, focused reads and list picking in about a second for about a cent.
BenchmarkbetaPythonCommunity directory of Jev apps, resources and runnable offline examples, with a guide skill and a searchable JevList web UI.
ArticleproductionPortuguese agent skill teaching Claude Code, Cursor and Codex to use Jev correctly, bundled with the full raw research behind each claim.
Agent skillproduction