User repository on GitHub
Fully customizable Advisor and Executor flow plugin for the Pi Coding Agent
- Forks
- 16
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.Homepage
CommunityJev Agents & MCP
Pi Advisor
Pi extension that keeps a fast Executor model on implementation and brings in a separate Advisor model for consequential choices and reviews.
- Stars
- 44
- Language
- TypeScript
- License
- MIT
- Published by
- Community
- Author
- philipbrembeck
- Last verified
- Oct 2026
About this project
A configurable second-opinion workflow for Pi coding agents, inspired by the "Steering Black-Box LLMs with Advisor Models" paper and Claude's Advisor feature. pi-advisor-flow keeps one model focused on execution and makes a second, smarter model available for consequential decisions, stalled work, and final reviews. The Executor still owns the work. The Advisor challenges assumptions, exposes risks, and suggests verification steps without taking over or running tools.
Quoted from the project's README. Read the full README.
Works forJev for Classification
Used inJev for Software & DevTools
- Languages
- TypeScript
Highlights
- On-demand ask_advisor reviews plus gates before plans, after repeated failures, and before completion with loop detection.
- Separate Executor and Advisor model controls with same-model skip, fallback retry, cached follow-ups, and exact-model whitelist.
- Per-response token and cost details folded into Pi cost totals plus advisor-stats adoption reporting and redaction controls.
- Optional Jev screening filter and proactive turn gate that skip low-stakes consultations; both stay off by default.
Watch out
Requires Pi 1.0.0 or later with configured models; same-model calls skip by default and follow-ups expire after minutes or new context. Copy reviewed against the repository README on 2026-10-04, not from running it.
Questions about this entry
Short answers drawn from the linked project or page.
Regular consultations never block execution and only run when asked or gated; same-model calls skip by default, Jev screening can skip low-stakes calls, and usage is normalized into Pi cost totals with per-response details.
The Executor keeps ownership and forms its own direction first, then the Advisor returns risks, alternatives, and verification steps without running tools. The Executor decides what to adopt and validates it in code.
An optional fallback model retries provider, auth, and availability failures once without consuming extra budget, and invalid advisor.json values fail only that call with a message instead of blocking tools.
It reconstructs only allowed conversation and repository context with secret redaction, trusted AGENTS.md rules, and explicit file handoff. Draft text stays untrusted and does not inherit tool policies, and visual reviews apply only where supported.
It requires Pi 1.0.0 or newer: run pi install npm:pi-advisor-flow, then reload Pi. /advisor enables the flow and /advisor-models picks the Executor, Advisor, and optional fallback. On first use, or when a saved model is unavailable, it opens the model picker rather than silently choosing one.
The Jev provider row in /advisor-settings chooses TypeSafe, an existing OpenRouter login, OpenAI Decisions (gpt-6-luna through advisorJevTransport openai-decisions), or any System One-compatible endpoint set with advisorJevBaseUrl. Every provider is live-verified before saving, and auto never picks the custom endpoint.
Calls and automatic gates are skipped with a notice by default, to avoid redundant consultations. Turn off Disable same-model Advisor in /advisor-settings, or set advisorDisableSameModel to false globally, if you intentionally want a higher-effort review from that same model.
More Jev resources like this
48GitHub stars
Agent harness for work that outlives one chat, with resumable turns, routed state machines, and observational memory across sessions and processes.
Jev Agents & MCP#community#jevbest#catalog-import
Community
dzhng/duet-agentTypeScript
dzhng/duet-agentTypeScript37GitHub stars
Assistant framework that stores specs, memory, and specialist agents as plain markdown files in your repo for Claude Code, Codex, Opencode, and similar runtimes.
Jev Agents & MCP#community#jevbest#catalog-import
36GitHub stars
Obsidian plugin providing multi-provider chat, vault read and write with function calling, visual workflow automation, and local vector search.
Jev Agents & MCP#community#jevbest#catalog-import
Community
takeshy/obsidian-llm-hubTypeScript
takeshy/obsidian-llm-hubTypeScript36GitHub stars
Pi extension that strips the full Agent Skills catalog from every request and loads only the skills a Jev ranking call judges relevant.
Jev Agents & MCP#community#jevbest#catalog-import
102GitHub stars
Behavioral code search that asks Jev to score repository fragments for a plain-language question and returns exact excerpts with paths and lines.
Jev Agents & MCP#community#jevbest#catalog-import
Community
nassim-arifette/jevgrepTypeScript
nassim-arifette/jevgrepTypeScript18GitHub stars
Local engine with TypeScript and Python SDKs that runs Claude Code and Codex sessions as a warm team with a durable mailbox.
Jev Agents & MCP#community#jevbest#catalog-import
Community
masonlee39/Multi-AgentTypeScript
masonlee39/Multi-AgentTypeScriptFrom the Jev community
Posts from builders shipping with Jev right now.