# Complex Feature Dev > Full-cycle 7-phase feature development workflow with persistent file-based planning (task_plan.md, findings.md, progress.md) as macro memory. Use when this capability is needed. - Skill: `tomevault-io/complex-feature-dev` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/complex-feature-dev` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/complex-feature-dev/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Productivity - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/complex-feature-dev --- # Feature Development (Complex Tasks) A codebase-agnostic workflow for building new features safely: - 7 phases (discovery → exploration → questions → architecture → implement → review → summary) - Persistent planning files for long tasks (macro memory: goal, status, next actions, and durable decisions) ## Inputs Minimum input: - A 1–2 sentence feature description (what to build + who/why). Helpful extras (optional): - Acceptance criteria / examples - Related files or modules - Docs / tickets / prototypes (if inaccessible, ask user to paste key parts) ## Non-Negotiable Rules 1. **Never skip Phase 3 (Clarifying Questions).** If anything is underspecified, ask and **wait**. 2. **Never start Phase 5 (Implementation) without explicit approval.** 3. **Use the planning files as persistent macro memory.** Keep them concise and durable. 4. **Read before decide.** Before major decisions, re-read `task_plan.md`. 5. **Log errors briefly + don’t repeat failures.** If an action fails, change the approach. ## Quick Start (Planning Files) This workflow requires these files in the **repo root**: - `task_plan.md` - `findings.md` - `progress.md` - `CLAUDE.md` (agent instructions; created by init if missing) To initialize them, run: - `/complex-feature-dev:init` (recommended) Or via terminal: - macOS/Linux (or Windows Git Bash): `bash "$(ls -dt ~/.claude/plugins/cache/complex-dev-skills/complex-feature-dev/*/scripts/init-session.sh 2>/dev/null | head -1)"` - Windows PowerShell: `pwsh -ExecutionPolicy Bypass -File "~/.claude/plugins/cache/complex-dev-skills/complex-feature-dev//scripts/init-session.ps1"` ## Project Guidelines - Follow `AGENTS.md` / `CLAUDE.md` / `CONTRIBUTING.md` / `README.md` if present. - Prefer matching existing code patterns over introducing new abstractions. - Write important conventions you discover into `findings.md`. --- ## Phase 1: Discovery **Goal:** Make the request concrete and testable. - If planning files are missing, run the initializer script via the Bash tool and do not proceed until the files exist. - Restate the request as acceptance criteria; confirm scope and non-goals. - Capture constraints (compatibility, performance, time, rollout). - Write brief confirmed requirements + acceptance criteria to `findings.md`. ## Phase 2: Codebase Exploration **Goal:** Identify the correct integration points and existing patterns. Do 2–3 independent exploration passes: - Pass A: find similar features and trace end-to-end - Pass B: map architecture boundaries and extension points - Pass C: identify testing/build/lint/config/observability conventions For each pass: - record key files/patterns (prefer file:line; keep it macro) in `findings.md`. Also: - locate and follow `AGENTS.md` instructions (if present). ## Phase 3: Clarifying Questions (Hard Stop) **Goal:** Resolve all ambiguity before architecture and implementation. Ask questions grouped by: - behavior definition (inputs/outputs/state) - edge cases (empty values, duplicates, concurrency, ordering, permissions) - error handling (timeouts, retries, idempotency, rollback) - integrations (APIs, data models, config, feature flags) - backward compatibility & migration - performance assumptions - observability - testing & acceptance plan Wait for answers (or propose defaults and get explicit confirmation). Record final answers in `findings.md`. ## Phase 4: Architecture Design **Goal:** Present 2–3 viable approaches and let the user choose. Provide at least: - Minimal changes (reuse, low risk) - Pragmatic balance (cleaner boundaries) - Optional clean architecture (more refactor, higher long-term clarity) For each approach include: - files to create/modify - key abstractions + data flow - trade-offs (risk/time/testability/maintainability) Write the chosen approach + rationale (brief) to `findings.md` and `task_plan.md`. ## Phase 5: Implementation (Requires Approval) **Goal:** Implement the chosen approach. - Implement in small verifiable increments. - Follow existing conventions; avoid unrelated refactors. - Log key actions/files/tests in `progress.md` (brief). - Log errors in `task_plan.md` (and brief resolution in `progress.md`). ## Phase 6: Quality Review **Goal:** Catch high-impact issues before delivery. Default review scope: - Prefer reviewing `git diff` (or specified files). Focus areas: - correctness, security/permissions, error handling, observability, conventions, simplicity Confidence filtering: - score issues 0–100 - report only issues ≥ 80 confidence, with concrete fixes (file:line) ## Phase 7: Summary **Goal:** Deliver a clean handoff. - Summarize what was built, key decisions, and files modified. - Provide verification steps (commands/endpoints/UI paths). - Call out risks and suggested next steps. --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/littlestudent886) — claim your Tome and manage your conversions.