agentic-development-workflow

Agentic Engineering Patterns

A Claude Code plugin for building software products with AI agents — from raw idea to shipped MVP.

Why This Exists

Traditional software development bottlenecks on human coding time. Process design optimizes "how to make people write code faster."

When agents can execute dozens of tasks in parallel, that bottleneck vanishes. A new one takes its place:

Agent execution capacity is near-infinite. Specification quality is not.

Vague specs don't slow down a human — they ask a colleague and adjust. Vague specs paralyze agents — they guess, diverge, and produce incompatible code across parallel sessions. The cost of ambiguity scales with parallelism.

This inverts the entire design logic:

Traditional:    plan roughly → adjust as you go → ship
                (optimizes for human coding speed)

Agentic:        invest heavily in spec precision → parallel execution → ship
                (optimizes for agent execution quality)

Every skill in this plugin serves that logic. The time you spend in /aep-envision and /aep-map pays back exponentially when agents build in parallel without asking questions.

Installing Skills

AEP skills follow the open Agent Skills format, so any project — under Claude Code, Codex, Cursor, OpenCode, and 70+ other agents — can install them with the skills CLI. No clone, no copied scripts.

Agent prompt

Prefer to delegate the install? Paste this to your coding agent — it covers the AEP skills, the required OpenSpec CLI, the formatter fix, and wiring AEP into your AGENTS.md / CLAUDE.md:

Install the Agentic Engineering Patterns (AEP) skills into this project, pinned to the latest release.

1. Install the latest release for each agent this repo uses — find the newest tag at
   https://github.com/memorysaver/agentic-engineering-patterns/releases/latest, then run once per agent:
     npx skills add memorysaver/agentic-engineering-patterns@<latest-tag> -a claude-code --skill '*' -y
     npx skills add memorysaver/agentic-engineering-patterns@<latest-tag> -a codex        --skill '*' -y
   This writes the skills under .claude/skills/ and/or .agents/skills/ plus a skills-lock.json manifest.

2. Commit the installed skill files together with skills-lock.json. The lockfile pins content
   hashes, not the git tag, so the committed files are what durably lock that release — after that,
   teammates and CI need no install step.

3. If this repo auto-formats Markdown/JSON on commit (Prettier, oxfmt, Biome, dprint, a
   lefthook/husky hook): exclude .claude/skills/**, .agents/skills/**, and skills-lock.json from
   the formatter, then make the install commit with --no-verify. Otherwise reformatting rewrites
   the skill files and breaks the lockfile hashes.

4. Add a short section to AGENTS.md (and/or CLAUDE.md) so the workflow is discoverable:
     ## AEP Workflow
     This project uses the Agentic Engineering Patterns (AEP) skills — a spec-driven, multi-agent
     feature lifecycle in .claude/skills/ and/or .agents/skills/, pinned via skills-lock.json.
     The skills are self-describing; start with `aep-onboard`. Upgrade by re-running
     `npx skills add memorysaver/agentic-engineering-patterns@<newtag>` once per agent.

5. Verify with `npx skills list`. Restore from the lockfile later with `npx skills experimental_install`.

6. REQUIRED — install the OpenSpec CLI. AEP is a spec-driven workflow: its skills (/aep-scaffold,
   /aep-dispatch, /aep-design, /aep-build, /aep-wrap) shell out to `openspec`, so it must be on PATH. Install it
   globally (needs Node >= 20.19), then verify:
     npm install -g @fission-ai/openspec@latest
     openspec --version
   AEP creates the per-project openspec/ artifacts itself — /aep-scaffold initializes them for a new
   project, or run `openspec init` once in an existing repo.