Mamdouh Aboammar is a vibe coder and conversion copywriter who builds agentic tooling for marketing work.
Agent governance, frontend quality gates, PR-only coding workflows, and Arabic conversion copy systems.
Every repository here started as a manual process that wasted enough of my week to be worth automating.
I don't want to become a software engineer
I want to become exceptionally good at deciding what technology should be built, why it should exist, how it should work, and whether it is actually succeeding
Career path
Founder of PrePilot — conversion infrastructure for Arabic-speaking markets. Digital Director with a background in performance marketing, Meta and Google campaign architecture, and direct-response systems for the MENA region. Earlier work included large-scale event marketing at the Hajj Conference and Exhibition. Now building open-source agentic tooling that makes the manual version of every workflow unnecessary.
|
Memory, trust boundaries, and failure lessons for Claude Code, Codex, Cursor, and Gemini CLI. Keeps an agent inside the rules you set.
|
23 design gates that run before an agent writes UI code. Generic layouts get blocked at the plan stage, well before review.
|
|
One CLI that routes bounded coding work across local agents, with a curated integration catalog and a full shipping chain.
|
Turns ChatGPT Web into a PR-only GitHub agent. It inspects a repo, implements one focused change, opens a reviewable PR, and stops there.
|
|
Research, drafting, and copy-audit pipeline tuned for MENA markets and Arabic conversion work.
|
Field notes, prompts, and routes for building vibe-coded apps with less guesswork and more reusable parts.
|
|
Runs a hardening playbook you own, validating each step and keeping a rollback path open.
|
Small animated 2.5D assets for developer docs and READMEs. Accessible by default, no build step.
|
|
Autonomous execution & work modeling engine for AI agents. Converts fuzzy ideas into verifiable deliverables with phase planning, evidence gates, and parallel workstreams.
|
Evidence gates and session continuity for AI coding agents. No evidence, no ship.
|
|
100+ Autonomous Skills & Specialized Division agents for Google Antigravity and Gemini CLI, covering engineering, design, security, and marketing.
|
Universal Cross-Host Agent Skill Compiler for ChatGPT, Codex, Claude Code, and Antigravity. Write a skill once, deploy it everywhere.
|
|
Make the model you already use work more like a frontier model — better planning, persistent context, skills, hooks, failure handling, and verification.
|
Temporal freshness guard for LLM agents, coding clients, MCP servers, and API middleware. Prevents stale reasoning from reaching production.
|
|
AI Agent skill for Multi-Agent Swarm Orchestration with budget caps, minimal fixes, and continuous Dev-QA gates.
|
Model-agnostic AI supervisor and deterministic permission engine for WhatsApp Cloud API.
|
|
Self-hosting ChatGPT/Codex Plugin autopilot for validation, deterministic packaging, and release verification.
|
Cross-agent marketing strategy pack with 29 Agent Skills, 24 specialist agents, neural marketing graph, and evidence challenge gates.
|
|
49+ AI agent skills & 51 CLI execution tools for technical marketers, founders, and AI coding agents.
|
Copy quality scoring engine. Rates persuasiveness, clarity, and conversion potential before copy ships.
|
|
Elite conversion copywriting & direct-response agent skill for Arab markets with AI anti-cliché filters.
|
Senior CRO & funnel auditor skill for landing pages, bundle offers, and A/B split-testing.
|
|
Reverse-engineered viral post generation for LinkedIn & Facebook with psychological hook triggers and zero-friction CTAs.
|
End-to-end tripwire & sample-to-paid funnel architecture skill for digital products and Meta CAPI.
|
|
Multi-format LinkedIn editorial system with visual layout prompts, hooks bank, and bilingual copywriting engines.
|
Claude Skill & CLI pipeline for deterministic, looping GIF infographics for LinkedIn. RTL & Arabic supported.
|
|
Influencer video scriptwriting skill in Gulf & Egyptian dialects for crisis management and PR defense.
|
Commercial Art Direction & Design Neural Mesh with 21 skills, dramaturgy-first AI video directing, and model-physics image direction.
|
|
Design system and pattern library for consistent, scalable UI across agent-generated interfaces and marketing surfaces.
|
Brand voice mirroring and content style extraction. Analyzes existing copy and reconstructs the author's style for AI-consistent output.
|
Problem before product The starting point is repeated friction, never a tool category. |
Context before automation The system knows the audience, the repo, the rules, and the evidence before it acts. |
Proof before claims A test, a report, a PR, or a measured result carries the conclusion. |
Judgment stays close Agents prepare, inspect, and verify. People own the consequential calls. |
| 3,931 Commit contributions |
61 days Current streak |
61 days Longest streak |
6,656 Contributions this year |
Pushonomics: Every line has a token tab
I keep a tab on what agent-assisted building costs, because the bill is part of the engineering decision.
Pushonomics: Every line has a token tab
I keep a tab on what agent-assisted building costs, because the bill is part of the engineering decision.
| 9,333,144 Lines added |
1,604,198 Lines deleted |
10,937,342 Lines changed |
349,994,944 Estimated tokens |
| GPT-5.6 Sol / Ultra $3,937.44 $5.00/M input · $30.00/M output |
Claude Fable 5 $6,999.90 $10.00/M input · $50.00/M output |
Estimate model: 2,766 commits scanned · 98 merges excluded · 262,496,208 input + 87,498,736 output tokens · 8 tokens per changed line · 4x session factor · estimate, not an invoice
