Local SQLite-based planning CLI for AI‑agent engineering tasks
Papertiger solves the problem of losing planning context between AI‑agent runs by storing tasks, dependencies, and decisions in a local SQLite database. It provides commands to add, query, and enforce task rules, ensuring that blockers and proof gates are respected before completion. Designed for developers integrating AI agents into their workflow, it runs as a lightweight Rust binary without needing a server. Its built‑in history and export/import features make it more reliable than ad‑hoc note‑taking or external issue trackers for long‑running engineering plans.
View on GitHub →virtualonno/papertiger