Creative Forge: agentic ad creation with deterministic validation and safe publishing
Creative Forge is an open-source pipeline for creating ad creatives using AI agents (Claude, Codex) with deterministic validators that enforce provenance, rights, localization, hashes, safe zones, timing, and exact artifact binding. It solves the problem of AI agents faking proof of successful publishing by requiring a fresh live readback showing the exact ad in PAUSED state before activation, ensuring ads are safe and correctly published.
Visit project ↗Entities
Related
AI-powered freelancer proposal tool built via AI-driven problem discovery
A tool that helps freelancers write better proposals by using a multi-step AI framework to identify and solve real freelancing problems. The creator used a 13-16 step process with multiple AIs to discover the problem and build a prompt-based solution, without personal freelancing experience.
Runeward: Sandbox AI agents with policy gates
Runeward is a sandboxing tool for AI agents that enforces policy gates to restrict agent actions. It uses LLMs to interpret and enforce user-defined policies, solving the problem of unsafe or unintended agent behavior for developers building autonomous AI systems.
Spendict: AI agent for performance marketing verdicts via MCP
Spendict is an AI agent that provides verdicts for performance marketers, using the Model Context Protocol (MCP) to integrate with marketing tools. It helps marketers evaluate campaign performance and make data-driven decisions.
Agent Legibility Analyzer: test if AI shopping agents can parse your store
A tool that lets e-commerce store owners check whether AI shopping agents can read and understand their product listings. It uses LLMs to simulate how agents interpret store content, helping merchants optimize their sites for agent-driven shopping.
Google’s SynthID embeds digital watermark during AI generation
Google’s SynthID takes a novel approach to AI content detection by embedding a digital watermark directly into generated content during creation, rather than analyzing it after the fact. This shift from post-hoc detection to proactive identification changes the architecture of AI content provenance.