wrannaman/agentic-engineering

The AX Stack — Agent Experience for engineering teams. From zero to 90% one-shot accuracy in 30 days. Skills, knowledge base, compound learning loop.

2Stars on the repository
21Mods indexed here, across every type
4mo agoLast push, which is what freshness is scored on
MITLicence, which decides whether bodies are shown

wrannaman/agentic-engineering

Skill Claude CodeCodex

Use this for adversarial design critique before implementation when the goal is to find flaws, poke holes, and identify edge cases, failure modes, abuse scenarios, and hidden assumptions. Stress-test ideas through focused dialogue, then transition to plan.

2 4mo ago A 54 tokens original MIT

gh-summary

02

wrannaman/agentic-engineering

Skill Claude CodeCodex

Summarize your merged PRs across repos and interpret the value delivered. Use when user wants to see what they've accomplished, prepare for standups, or reflect on their contributions.

2 4mo ago A 39 tokens original MIT

spec

03

wrannaman/agentic-engineering

Skill Claude CodeCodex

Write design specs and architecture decision records (ADRs) with deep research and gotcha hunting.

2 4mo ago A 21 tokens original MIT

brainstorm

04

wrannaman/agentic-engineering

Skill Claude CodeCodex

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

2 4mo ago A 36 tokens copy · 73% MIT

compound

05

wrannaman/agentic-engineering

Skill Claude CodeCodex

Extract learnings from current session, search past sessions with CASS, and document patterns to prevent recurring mistakes.

2 4mo ago A 24 tokens original MIT

plan

06

wrannaman/agentic-engineering

Skill Claude CodeCodex

Pattern lookup, research, and design phase. NO coding allowed - only research and planning.

2 4mo ago A 20 tokens original MIT

review

07

wrannaman/agentic-engineering

Skill Claude CodeCodex

Multi-perspective code review with 8 parallel passes. Language-agnostic — adapts to whatever your codebase uses via KB.

2 4mo ago A 29 tokens original MIT

seed

08

wrannaman/agentic-engineering

Skill Claude CodeCodex

The #1 reason agentic engineering fails is an empty knowledge base. Engineers are told "write docs for the KB" and either procrastinate or write docs so generic they're useless.

2 4mo ago A 0 tokens original MIT

setup

09

wrannaman/agentic-engineering

Skill Claude CodeCodex

Interactive guided setup. Detects your codebase, configures MCP, writes project-specific CLAUDE.md, seeds the KB, and verifies everything works.

2 4mo ago A 33 tokens original MIT

ship

10

wrannaman/agentic-engineering

Skill Claude CodeCodex

End-to-end: plan → implement → review → fix → push in one shot. No human feedback between steps. Use when you trust the system and just want the thing done.

2 4mo ago A 38 tokens original MIT

work

11

wrannaman/agentic-engineering

Skill Claude CodeCodex

Execute implementation from a plan with verification at each step.

2 4mo ago A 13 tokens original MIT

pr-push

13

wrannaman/agentic-engineering

Skill Claude CodeCodex

Create or update a draft PR, auto-fixing pre-commit failures until push succeeds.

2 4mo ago A 22 tokens original MIT

rebase-fix

14

wrannaman/agentic-engineering

Skill Claude CodeCodex

Resolve merge/rebase conflicts and continue restacking until the entire stack is clean. Use when the user says "fix rebase", "fix conflicts", "resolve conflicts", "restack", "gt sr", "stack rebase", or when a rebase or restack operation has stopped due to merge conflicts.

2 4mo ago A 68 tokens original MIT

stack-pr

15

wrannaman/agentic-engineering

Skill Claude CodeCodex

Large features split into small, reviewable PRs that build on each other. This skill creates the stack: branches, draft PRs, and dependency links. Uses plain git + gh CLI — no special tooling required.

2 4mo ago A 0 tokens original MIT

deslop-code

17

wrannaman/agentic-engineering

Skill Claude CodeCodex

Aggressively reduce and clean up code in any language — strip AI slop, remove bloat, collapse indirection, and shrink files to their essence. ALWAYS use this skill when the user wants code made shorter, leaner, or less verbose — whether they say "deslop", "reduce", "simplify", "trim", "slim down", "too much code"…

2 4mo ago A 143 tokens original MIT

simplify

19

wrannaman/agentic-engineering

Skill Claude CodeCodex

After implementation, take a second pass looking for opportunities to simplify. Not the same as deslop (which removes AI bloat) — this is about finding cleaner approaches to what you just built.

2 4mo ago A 0 tokens original MIT