Loading catalog...
This website and its associated Model Context Protocol (MCP) server are operated as an independent, non-commercial academic research initiative to advance public understanding of autonomous agent architectures and interface protocols.
Under 17 U.S.C. § 107 (United States) and Directive (EU) 2019/790 (European Union) on text and data mining for scientific research, the extraction, synthesis, and conceptual indexing of publicly scheduled technical conference metadata constitutes non-infringing transformative fair use. All distillations are original AI-assisted analytical syntheses rather than verbatim transcriptions.
This platform does not store, host, mirror, or redistribute speaker slide PDFs, presentation files, or proprietary audiovisual recordings. All citations, speaker attributions, and deck references route strictly and directly to canonical, author-hosted URLs on the official conference schedule.
"AGNTCon", "MCPCon", "Linux Foundation", "Sched", and associated logos or marks are registered trademarks of their respective owners. This project is an independent community open-source effort and is not endorsed by, sponsored by, or affiliated with The Linux Foundation or Sched.com.
We deeply respect speaker intellectual property and community goodwill. If you are a speaker or rights holder and wish to update your talk's essence, append verified code repositories, or request exclusion/takedown, please submit our verified speaker form below.
This website is an independent open-source research and engineering demonstration developed for the global AI and developer community.
This project is created independently and is not affiliated with, endorsed by, or sponsored by The Linux Foundation, Sched.com, or event sponsors.
This demonstration is provided as-is for research and exploration. The author does not operate it as a commercial SaaS product and does not guarantee continuous 24/7 uptime or ongoing commercial service level agreements (SLAs).
We deeply respect intellectual property and speaker goodwill. If you are an author or speaker and wish to propose improvements, append verified code repositories, or request exclusion, please submit our verified form.
• Standard Laptops (8–16 GB RAM): ollama run qwen2.5:7b-instruct or llama3.1:8b
• Low Resource (4–8 GB RAM): ollama run llama3.2:3b-instruct or qwen2.5:3b
• Browser Access Flag: Ensure Ollama runs with OLLAMA_ORIGINS="*" ollama serve to enable browser requests.
Download pre-indexed bundles for instant offline search in Cursor, Claude Desktop, and OpenClaw without live API calls:
Includes: Pre-indexed SQLite FTS5 (BM25 Lexical Full-Text Search — Zero Local Embedding Dependencies) and pure agntcon2026.jsonl.
For complete step-by-step setup guides in Cursor, Claude Desktop, VS Code (Roo Code / Cline), and OpenClaw, see the AGNT_MCP_con_2026 Documentation ↗.
{
"mcpServers": {
"agntcon2026": {
"command": "python",
"args": ["-m", "mcp_server", "--data", "wiki/agntcon2026.sqlite"],
"env": {}
}
}
}Tailor session relevance scores, assistant answers, and recommendations to your technical goals. Zero Server PII Guarantee: Profile data is stored 100% in your browser's local memory.
Curate your personal conference itinerary. Add sessions from the live search picker or auto-curate from your profile, then export to Obsidian or PDF.