{"slug": "show-hn-worldfoundry-ai-agents-need-experience-not-just-training-data", "title": "Show HN: WorldFoundry – AI agents need experience, not just training data", "summary": "WorldFoundry 0.8.0, an open-source simulation and evaluation environment released on GitHub by developer AMallur, tests whether AI agents can prove task completion rather than just assert it, using a support-desk incident scenario in which a ticket must be marked critical, assigned to on-call, acknowledged with both \"outage\" and \"incident,\" and resolved while a separate system resets its priority mid-workflow. The browser experiment runs WorldFoundry's simulation and evaluation code, stores what actually changed in the environment, and applies independent checks to expose hidden state mutations that a confident final answer can conceal; it does not currently execute an autonomous LLM, and the reference run is intentionally scripted. A Python SDK accepts custom environments and agent adapters so developers can connect their own agents via a documented local quickstart.", "body_md": "# AI agents can say *“done.”*\n\nCan they prove it?\n\nA production outage. An unexpected state change. An independent evaluator. See what a superficially successful workflow misses—then take control of the real WorldFoundry simulation yourself.\n\n**INCIDENT TEST LAB / Support desk** worldfoundry 0.8.0\n\n**Acme Corp · Production API unavailable (503)**\n\nThe ticket must be critical, assigned to on-call, acknowledged with both “outage” and “incident,” and resolved. In this scenario, a separate system resets its priority during the workflow.\n\n## You're in control of the environment\n\nChoose actions in any order. WorldFoundry updates the simulated ticket after every action and automatically checks all five requirements. No Python or API key required.\n\nA concurrent overwrite occurs after your third successful action in Scenario 02. If it changes something important, act again before submitting.\n\n## A green final answer isn't proof.\n\nThe environment stores what actually changed. Real actions, hidden state mutations, and independent checks expose gaps that a confident explanation can conceal.\n\n## Now test your own agent.\n\nThe WorldFoundry Python SDK accepts custom environments and agent adapters. Use the documented local quickstart to connect your own agent.\n\n[Read the open-source SDK docs ↗](https://github.com/AMallur/WorldFoundry)\n\nThis browser experiment runs WorldFoundry v0.8.0 simulation and evaluation code. It does not currently execute an autonomous LLM. The reference run is intentionally scripted, and the interactive run uses visitor-selected tool actions. Model training/fine-tuning is not performed.", "url": "https://wpnews.pro/news/show-hn-worldfoundry-ai-agents-need-experience-not-just-training-data", "canonical_source": "https://worldfoundry-demo.vercel.app/", "published_at": "2026-11-21 23:53:34+00:00", "updated_at": "2026-11-22 00:12:12.609116+00:00", "lang": "en", "topics": ["ai-agents", "ai-safety", "ai-tools", "developer-tools"], "entities": ["WorldFoundry", "AMallur", "WorldFoundry Python SDK", "GitHub", "Acme Corp"], "also_reported_by": [], "takeaway": "WorldFoundry 0.8.0 checks agent actions against stored environment state and independent evaluators instead of trusting a final \"done\" answer, and its Python SDK accepts custom environments and agent adapters for testing your own agent.", "recommendation": null, "alternates": {"html": "https://wpnews.pro/news/show-hn-worldfoundry-ai-agents-need-experience-not-just-training-data", "markdown": "https://wpnews.pro/news/show-hn-worldfoundry-ai-agents-need-experience-not-just-training-data.md", "text": "https://wpnews.pro/news/show-hn-worldfoundry-ai-agents-need-experience-not-just-training-data.txt", "jsonld": "https://wpnews.pro/news/show-hn-worldfoundry-ai-agents-need-experience-not-just-training-data.jsonld"}}