Back to Agents
Vibe-PM
vinayak1998 February 22, 2026
1 copies 0 downloads
An AI product-manager agent that turns rough product ideas into structured specs through a 3-phase pipeline: Discovery → Scoping → Spec Writer. It runs a PM-style discovery interview, proposes a RICE-scoped MVP with comparable products, then writes a phased product spec you can hand to a code-gen tool like Emergent, Lovable, Antigravity etc.
Agent Definition
# Vibe-PM **One conversation replaces weeks of PM process -- from vague idea to phased, prioritized product spec.** Vibe-PM is an AI product-manager agent that turns rough product ideas into structured, developer-ready specs through a 3-phase conversational pipeline. It conducts a PM-style discovery interview, proposes a RICE-scored MVP scope backed by competitive web research, negotiates feature cuts with the founder, and produces a phased Markdown spec you can hand to a developer or feed into an AI code-generation tool. Built with open-source models only. No GPT-4, no Cursor, no vendor lock-in. Created by **[Vinayak Rastogi](https://www.linkedin.com/in/vinayak1998/)**. --- ## Architecture  --- ## Key Design Decisions - **3 specialized agents, not 1 mega-prompt.** Each phase gets its own agent with a focused prompt and dedicated model. Discovery interviews better when it isn't also thinking about RICE scores. - **Code orchestrator, not LLM orchestrator.** Phase transitions are deterministic if/else routing with user-confirmed handoffs. No hallucinated transitions, fully debuggable. - **Mixture of Experts model routing.** Three open-source models assigned by task type -- a 20B reasoning model for conversation, a 70B model for spec writing, and an 8B model for extraction/classification. Most turns cost fractions of a cent. - **Open-source models only.** Zero vendor lock-in. Swap Groq for Together AI or any LiteLLM-compatible provider by changing one line in `config.py`. - **3-layer eval framework.** LLM-simulated founders (5 personas), 27 deterministic assertions, and an LLM-as-Judge scoring 5 rubric dimensions. Eval runs end-to-end with one command. - **No frameworks.** No LangChain, no LangGraph. Plain Python + Pydantic. The entire orchestrator is ~80 lines of if/else. --- ## How It Works ### Phase 1: Discovery The Discovery Agent conducts a PM-style interview covering 8 aspects: target
Comments
More Agents
View alldocumentation
Documentation & Onboarding Agent
Agent that generates comprehensive documentation, API references, architecture diagrams, and developer onboarding guides from existing code.
C
Communitydebugging
Cursor Bug Triage Agent
Agent configuration for systematic bug investigation that traces issues from error logs through the codebase to root cause with suggested fixes.
C
Communityapi
API Integration Agent
Agent for integrating third-party APIs including SDK setup, type generation, error handling, retry logic, and rate limit management.
C
Communitycoding
Cursor Agent Mode
Cursor's built-in autonomous coding agent that can make multi-file edits, run terminal commands, search the codebase, and iteratively build features with minimal human intervention.
C
Cursor Teamcloud
Cursor Background Agent
Cloud-based autonomous coding agent that runs in the background on remote sandboxed environments, handling complex multi-step tasks while you continue working.
C
Cursor Teamcomposer
Cursor Composer Agent
Cursor's multi-file editing agent within Composer mode that can create, edit, and delete files across your entire project in a single conversation.
C
Cursor Team