LLM-powered AI Agents and the Semantic Web Symbiosis Showcase

Hi Everyone,

Agent context naturally informs agent behavior. But, like so many 
aspects of AI, this topic has become dominated by marketing narratives 
that overlook the power RDF brings to the conversation.

We've been working on changing that, as discussed in several of my 
recent posts across LinkedIn and X.

Our latest effort introduces RDF-based context for handling session 
identity, preferences, logs, how-tos, and more. Together, these provide 
machine-computable context that can guide any AI agent loosely coupled 
with a variety of LLMs through your preferred tooling, including 
Anthropic Claude Code and CoWork, OpenAI Codex, OpenCode, and others.

*Why should you care?*

Instead of relying on increasingly complex prompt engineering and fuzzy 
prose, your AI agent can leverage precise, machine-computable context 
that delivers more deterministic, consistent, and reusable behavior. The 
result is a better agent experience, regardless of the underlying LLM or 
tooling.

Fundamentally, this approach keeps LLMs focused on what they do 
best—processing fuzzy natural language—while RDF and SPARQL query 
templates handle deterministic interactions with Semantic Webs. As part 
of an agent harness loop, they collectively leverage the combined 
strengths of filesystems and database management systems that support 
open standards such as RDF, SPARQL, WebDAV, WebID/NetID, PKI, and many 
others.

*Links:*

[1] https://www.linkedin.com/in/kidehen/recent-activity/newsletter/

[2] 
https://linkeddata.uriburner.com/DAV/demos/daas/2026-07-16-session-context-engineering-showcase.html 
-- HTML with embedded screencast, now with a voice-over, generated using 
these capabilities via a collection of loosely coupled skills.

[3] https://github.com/OpenLinkSoftware/ai-agent-skills -- Agent Skills 
Github Repo that includes RDF-based Agent context management functionality

-- 
Regards,

Kingsley Idehen 
Founder & CEO
OpenLink Software
Home Page:http://www.openlinksw.com
Community Support:https://community.openlinksw.com

Social Media:
LinkedIn:http://www.linkedin.com/in/kidehen
Twitter :https://twitter.com/kidehen