Back to directory

Planning with Files

A file-based planning skill for AI coding agents that keeps a task plan, findings and progress on disk so long-running work survives context clearing, crashes and compaction.

Planning with Files

Planning with Files is a persistent, file-based planning skill for AI coding agents and long-running agent tasks. It keeps a task plan, a findings file and a progress file on disk instead of inside the context window, and lifecycle hooks re-inject the relevant planning context every turn, so the plan survives a context clear, a crash, or compaction.

Automatic recovery reads only the project's own plan files; reviewing past session records for aggregate counts or a bounded replay needs an explicit catch-up mode. The skill installs across more than 60 agents through the Agent Skills standard, with native plugins for Claude Code, Codex CLI, Pi, Hermes Agent, OpenCode and DeepSeek Harness.

In the project's own measurements, keeping the plan on disk reduced a 13.3-turn re-orientation after context loss down to 5.0 turns. It installs with a single command through the skills tool and is written mainly in Shell, under the MIT license.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.