RFC: Exploring CommerceTXT (v1.0.1) integration for AI-agent discovery #4
Description
Following our discussion on Hacker News, I've formalized the CommerceTXT specification to address deterministic commerce data delivery for AI agents.
What is CommerceTXT?
CommerceTXT is a lightweight, read-only protocol (v1.0.1) that lets stores expose pricing and inventory data in an agent-friendly format — without HTML scraping.
The goal is intentionally modest:
- Reduce bot/agent bandwidth by ~95%
- Provide hallucination-free pricing and inventory to LLM-based shoppers
- Complement existing HTML + JSON-LD setups, not replace them
Why this matters for Openfront merchants
- AI shopping agents (ChatGPT, Claude, Perplexity) currently can't discover Openfront stores efficiently
- Early positioning: Openfront merchants could be among the first AI-native storefronts
- Minimal infrastructure overhead: A view-layer mapping of existing product data to text
- Permissionless discovery: No API keys, no vendor lock-in — just a text file
What I'm proposing
I've listed Openfront as a first integration partner in our contributors list, following our discussion on Hacker News.
If useful, I'm happy to support the discussion with documentation, examples, or a small proof-of-concept at a later stage.
No expectations on timelines — just exploring fit and feasibility.
Resources
- Full specification: https://github.com/commercetxt/commercetxt/blob/main/spec/README.md
Would you be open to a short async discussion here to explore fit and feasibility?
Best,
Tsanko