User repository on GitHub
Bounded TypeSafe Jev workflows for coding agents.
- Forks
- 6
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Tools & CLIs
Stanley
A Jev-first command-line coding agent: natural-language requests are routed by Jev to one of ten bounded review and triage workflows, and unsupported requests fall back to the Pi coding agent or a workflow Stanley stages for promotion.
- Stars
- 117
- Language
- TypeScript
- License
- MIT
- Reading time
- 3 min
- Published by
- Community
- Author
- devagrawal09
- Last verified
- Sep 2026
About this project
Stanley is a command-line tool for checking and understanding code changes. Instead of naming a command, you describe what you need. TypeSafe Jev routes the request to one workflow: a built-in, a trusted repository workflow, or one Stanley wrote for itself. Each workflow is deterministic code that gathers bounded evidence and asks Jev small fixed-choice questions about it; code, not a model, makes the decisions. That is the hot path, and it never starts a general agent.
Quoted from the project's README. Read the full README.
Used inJev for Security & Privacy
Highlights
- Ten built-in workflows cover diff checks, security, performance and compatibility review, test-gap finding, and failure triage.
- Not on npm yet: 0.1.0 is built from source, and the old jev-code package on npm is 0.0.1 and does nothing.
- Unsupported requests go to the Pi coding agent when installed, then Stanley drafts a workflow it validates and stages for you to promote.
- The TypeSafe key is read only from TYPESAFE_API_KEY, never files or flags, and the model comes from TYPESAFE_MODEL.
- The README warns that an empty findings list is not an approval; reports also list notChecked.
Quickstart
bash
git clone https://github.com/devagrawal09/stanley-code.git
cd stanley-code
npm ci
npm run build
export TYPESAFE_API_KEY="YOUR_API_KEY"
node dist/cli.js --helpWatch out
MIT-licensed and not affiliated with TypeSafe; requires Node.js 22.18 or newer, git, a repository, and a TypeSafe API key, and Windows is untested.
Questions about this entry
Short answers drawn from the linked project or page.
Not on the hot path. Jev routes your natural-language request to one of ten bounded workflows made of deterministic code that gathers evidence and asks Jev fixed-choice questions, and no general agent starts. Only a request no workflow supports falls through to the Pi coding agent, if it is installed.
It stages them instead of installing them. An agent drafts a workflow for the request type Stanley could not handle, the candidate is validated and staged under .stanley/workflows/, and you promote it; only after that does that kind of request run without an agent.
Build it. stanley-code is not on npm; the README describes an unreleased 0.1.0 and tells you to clone, run npm ci, npm run build, then node dist/cli.js. The old jev-code package on npm is a 0.0.1 placeholder that does nothing.
No. The README states plainly that an empty findings list is not an approval, and reports also carry a notChecked section. Coverage is capped by policy - the first 150 hunks per analysis, the top 3000 files screened - so silence can mean unexamined rather than clean.
Jev chooses one built-in or repository workflow from author-provided JSON routing metadata, after deterministic facts such as whether a diff exists gate eligibility. A pick needs confidence at or above 0.6, probability at or above 0.55 and a margin at or above 0.15; low confidence, a narrow margin or an unavailable pick all return cannot_tell rather than a guess.
Find relevant code, check current changes, review current changes, find test gaps, summarize current changes, review security, review performance, review compatibility, triage test failures, and triage review comments. Trusted repository workflows under .stanley/workflows/ extend them, and anything else is delegated to the Pi agent when installed.
Until 0.1.0 is on npm, build from source: git clone, npm ci, npm run build, then run node dist/cli.js. Node.js 22.18 or newer, git and a Git repository are required. The key is read only from the TYPESAFE_API_KEY environment variable, never from files or flags. CI tests on Linux and Windows is untested.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
X: Decision outputs in Grep
Hacker News: jevmem keeps Claude Code project memory by asking Jev which turns are worth saving.
avinashjetwani61 pointsJevmem – automatic project memory for Claude Code, built on JevRead the thread on Hacker NewsX: A usage router for Grok Bot
X: The compaction plugin, in Chinese
More Jev resources like this
Agent-ergonomic CLI following the AXI conventions that gives coding agents Jev judgments for blocking risky tool calls, screening fetched content for prompt injection, triaging logs, flagging risky diffs, and ranking many items. Routine commands are decided locally at no cost.
Jev Tools & CLIs#community#typescript#cli
Community
shiftynick/jev-axiTypeScript
shiftynick/jev-axiTypeScriptPi extension that supervises a coding agent with Jev judgments: it holds risky tool calls, checks writes against project Markdown rules, and feeds most issues back to the agent as a steer instead of interrupting you. The conscience is beta and off by default.
Jev Tools & CLIs#community#typescript#guardrails
Community
DevMortimer/pi-wardenTypeScript
DevMortimer/pi-wardenTypeScriptRust coding-agent harness (MIT, ~20k stars) whose crates/jcode-base/src/jev.rs is a typed TypeSafe Jev decision transport, used for memory pruning (memory_jev.rs), browser navigation (browser_jev.rs), and voice routing (voice_intent.rs) with separate provider selectors per purpose.
Jev Tools & CLIs#community#rust#coding-agent
Community
1jehuang/jcodeRust
1jehuang/jcodeRustA CLI that asks Jev which files and declarations answer a repository question, then prints verbatim excerpts for a coding agent to use.
Jev Tools & CLIs#community#typescript#cli
Community
dzhng/jevgrepTypeScript
dzhng/jevgrepTypeScriptA collection of Jev demos, workflows, and agent skills for coding agents: five installable skills and 108 scenarios, installed by pointing Codex, Claude Code, or OpenCode at an agent prompt.
Jev Tools & CLIs#community#python#skills
Community
wuyoscar/jev-skillPython
wuyoscar/jev-skillPythonA staged code-review workflow that uses Jev for bounded judgments over Git diffs or whole codebases, with a local dashboard. Policy and thresholds stay in code while Jev screens five risk areas.
Jev Tools & CLIs#community#typescript#code-review
Communitydevagrawal09/jev-reviewTypeScript
devagrawal09/jev-reviewTypeScriptFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.