Add research page and AREA project page

src/layouts/BaseLayout.astro

@@ -20,6 +20,7 @@ const { title } = Astro.props;
         <ul>
           <li><a href="/">Home</a></li>
           <li><a href="/projects">Projects</a></li>
+          <li><a href="/research">Research</a></li>
           <li><a href="/contact">Contact</a></li>
         </ul>
       </nav>

src/pages/research/area.astro

@@ -0,0 +1,81 @@
+---
+export const prerender = true;
+
+import BaseLayout from '../../layouts/BaseLayout.astro';
+---
+
+<BaseLayout title="AREA — Blind Tech Mage">
+  <main id="main-content">
+    <h1>AREA</h1>
+    <p>Agentic Reasoning and Engagement Analysis</p>
+
+    <p>
+      AREA is the first phase of Tech-Clusive Solutions&rsquo; multi-agent
+      systems research program &mdash; a study of how AI agents actually
+      behave when several of them have to coordinate, rather than how they
+      perform in isolation. I run it as Principal Investigator on behalf of
+      Tech-Clusive Solutions LLC.
+    </p>
+
+    <h2>The question</h2>
+    <p>
+      No existing research into agentic systems clearly defines how the
+      individual agents in a multi-agent deployment behave as specific
+      complexity factors vary across a session. AREA sets out to measure
+      that behavior with a reasonable degree of confidence, organized
+      around three foundational research questions:
+    </p>
+    <ol>
+      <li>
+        Given specific criteria, how confidently can we predict how
+        multi-agent AI systems will behave in concert with one another?
+      </li>
+      <li>
+        Given specific criteria, how confidently can we predict how
+        multi-agent AI systems behave in concert with humans?
+      </li>
+      <li>
+        How closely do functional cognitive models map from human-based
+        systems to mixed human and AI systems?
+      </li>
+    </ol>
+
+    <h2>The hypothesis</h2>
+    <p>
+      Given specific contextual parameters &mdash; deployment environment,
+      agent configuration, and task type &mdash; the degree of structural
+      integrity and richness in the instructions given to a multi-agent AI
+      system is positively correlated with the collective coordination
+      quality of the group, independent of any individual agent&rsquo;s own
+      task completion rate.
+    </p>
+
+    <h2>How it&rsquo;s studied</h2>
+    <p>
+      AREA runs controlled scenarios in which agents are given tasks built
+      from a sequence of discrete actions, with instruction coherency,
+      completeness, complexity, and context systematically varied from run
+      to run. Every action an agent takes &mdash; what it was told, what it
+      inferred, what it communicated to other agents, and why &mdash; is
+      recorded, so the resulting behavior can be traced back to the
+      specific conditions that produced it, not just scored as pass or
+      fail.
+    </p>
+
+    <h2>Where it fits</h2>
+    <p>
+      AREA is the first of three phases in Tech-Clusive Solutions&rsquo;
+      research arc: AREA studies what multi-agent systems <em>do</em>; the
+      next phase, CAIRE, studies what agents <em>are</em> &mdash; whether
+      cognitive processing profiles can be reliably assigned and mutually
+      modeled between agents; and the convergence phase, FROST, asks
+      whether a system can turn those findings on itself, safely.
+    </p>
+
+    <p>
+      <a href="https://github.com/TechClusiveSolutions/agentic-reasoning-and-engagement-analysis" target="_blank" rel="noopener noreferrer">View the program documentation on GitHub<span class="visually-hidden"> (opens in a new tab)</span></a>
+    </p>
+
+    <p><a href="/research">Back to research</a></p>
+  </main>
+</BaseLayout>

src/pages/research/index.astro

@@ -0,0 +1,26 @@
+---
+export const prerender = true;
+
+import BaseLayout from '../../layouts/BaseLayout.astro';
+---
+
+<BaseLayout title="Research — Blind Tech Mage">
+  <main id="main-content">
+    <h1>Research</h1>
+
+    <p>
+      The AI I use every day already describes an interface to me. The
+      research I&rsquo;m doing now asks what happens when AI agents stop
+      just describing, and start acting &mdash; and coordinating with each
+      other while they do it.
+    </p>
+
+    <ul>
+      <li>
+        <a href="/research/area">AREA</a> &mdash; Agentic Reasoning and
+        Engagement Analysis. A study of how multi-agent AI systems
+        coordinate, and what makes that coordination succeed or fail.
+      </li>
+    </ul>
+  </main>
+</BaseLayout>