Menu
Plan implementation before writing code. Analyzes codebase structure and dependencies, asks clarifying questions, and writes a scoped implementation spec.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
| name | spec |
| description | Plan implementation before writing code. Analyzes codebase structure and dependencies, asks clarifying questions, and writes a scoped implementation spec. |
| argument-hint | <feature or change description> |
| allowed-tools | Bash Skill |
Before writing any code for: $ARGUMENTS
Step 1 — Structural analysis
Run /code-analysis to get a codebase briefing — top-level modules, structure, and key files. Then identify which modules and functions this change will touch. Batch them into a single /code-analysis call using comma-separated targets or graph operators — do not invoke /code-analysis once per target.
Use the results to understand:
Do not read files directly. Use the source blocks returned by the analysis instead.
Step 2 — Clarifying questions
Ask me about requirements, edge cases, and constraints. Use the structural analysis to ask informed questions — e.g. "the analysis shows X depends on Y, should the spec account for updating Y as well?"
Step 3 — Write the spec
Write a concise implementation plan to .claude/scratch/spec-[feature-name].md in the current directory. The spec should:
Step 4 — Wait for approval
Do not write any code until I approve the spec.
Bump the project version across all files that carry it. Pass the new version as the argument (e.g. /bump-version 0.4.0).
Full code navigation — replaces grep, glob, and file reads. Find symbol definitions, trace callers and callees, read source code, and map module structure. Pass symbol names or questions as arguments.
Generate a PR description as a Markdown file in .claude/scratch using git history and code analysis. Fills in the project's PR template.
Diagnose and fix a bug or unexpected behavior. Traces the problem through the dependency graph, adds logging if needed, and proposes a fix. Use when something is broken, producing wrong output, or behaving unexpectedly.
