storybloq

Why we included this project

Anyone who spends their day inside Claude Code or Codex knows the reset problem: every new session starts with the model having no memory of what was built yesterday, so you re-explain the project state before you can get back to work. Storybloq fixes that by keeping tickets, issues, handovers, and roadmap decisions in a git-tracked .story/ directory that the assistant loads at the start of each session. The project gives you three ways to work with that directory, a CLI for inspecting and editing it, an MCP server that exposes structured tools to the coding agent without spawning subprocesses, and a /story skill for in-session use, so you can pick the layer that fits your workflow. Because everything is plain JSON and markdown inside the repo, the context is versioned with your code and stays readable by any tool instead of being locked into a proprietary store. That makes it a practical fit for solo developers and small teams who want context that compounds across sessions without running a separate service.