Plnr
AI planning and security CLI for codebases that turns architecture context into concrete implementation plans.
Project Overview
Plnr is a CLI built for the phase before implementation starts: understanding the codebase, mapping the architecture, and turning ambiguity into an actionable plan. It inspects frameworks, dependencies, code patterns, and security risks, then combines that context with modern LLMs to produce implementation guidance that is grounded in the repo instead of generic advice.
I also built dedicated planning, chat, and security flows so it can answer architecture questions, generate step-by-step execution plans, and run OWASP-style audits from the terminal. The project is especially strong for backend-heavy repositories where context gathering and risk reduction matter before code gets written.
Technical Challenges
Fitting rich codebase context into model windows without losing the signals that matter for planning.
Building prompt flows that can switch between chat, planning, and security review while still staying context-aware.
Integrating live docs and web search through MCP so generated plans are based on current library behavior.
Designing a CLI UX that feels fast and readable even when analysis spans many files and multiple reasoning steps.
Tech Stack
Status
Current Status
Completed & Live
Last Update
2025
