from ed3d-plugins206
Generate detailed, engineer-ready implementation plans from design docs: file paths, step-by-step tasks, tests, and verification instructions.
This skill turns high-level design plans into actionable implementation plans for engineers who have no prior context on the codebase. It enforces scope checks, codebase verification via subagents, external dependency research, and generates granular tasks with absolute paths, acceptance-criteria mapping, and test requirements. The output is a set of phase files and a test-requirements.md ready for execution.
Use when a design document exists and you need a runnable implementation plan (≤8 phases) that includes codebase verification, dependency research, and per-phase tasks. It is intended for teams that want reproducible, reviewable plans and explicit verification steps before coding begins.
Best used by agent runtimes that support subagent dispatch and workspace file operations (Claude Code, ed3d tooling, advanced agent orchestrators).
A comprehensive prompt-only skill for generating detailed implementation plans from design docs. It provides structured templates, phase-by-phase workflows (interactive and batch modes), and an extensive anti-rationalization table to prevent common shortcuts. No bundled scripts to test. The skill is well-written but very long and heavily coupled to the ed3d ecosystem (requires codebase-investigator, internet-researcher, remote-code-researcher, code-reviewer subagents).
Purely instructional skill with no scripts. Security is clean — no network calls, no credentials, no destructive commands. The skill is impressively detailed with strong guardrails against common LLM planning shortcuts. The 8-phase limit and mandatory codebase verification are good design choices. Main limitation is ecosystem specificity and the verbosity of the SKILL.md itself.
