What problem does it solve? Multi-step feature work often fails because plans are vague, tasks are too large, and engineers lack context about the codebase. This Skill produces implementation plans detailed enough that an engineer with zero context can execute them without assumptions. ## Core Features & Use Cases - Zero-Context Plan Format: Every task specifies exact file paths, complete code, exact commands, and expected output. - Bite-Sized Task Granularity: Each task is a single 2-5 minute action following a TDD cycle of failing test, minimal implementation, verification, and commit. - Octopus Integration: Plans can be executed sequentially, parallelized via orchestrate.sh tangle, or run through the full embrace workflow, with session-aware plan storage. - Use Case: Given a spec for user authentication, produce a plan where each task shows the exact test file, the failing test code, the implementation code, the test command with expected output, and the git commit message. ## Quick Start Ask the AI to create a zero-context implementation plan for your multi-step feature before writing any code.