docs: cloud-native messaging pivot

.github/workflows/deploy.yml

@@ -8,9 +8,6 @@ on:
       - main
       - master
 
-concurrency:
-  group: ${{ github.workflow }}-${{ github.ref }}
-  cancel-in-progress: true
 
 jobs:
   Deploy-Production:

src/lib/data/agents.ts

@@ -5,6 +5,7 @@ export interface AgentData {
   name: string;
   icon: any;
   color: string;
+  cloudNativeRole: string;
   description: string;
   longDescription: string;
   githubLink: string;
@@ -18,7 +19,8 @@ export const agents: AgentData[] = [
     name: "Agent Body",
     icon: Server,
     color: "from-slate-500 to-slate-700",
-    description: "The root supervisor. Daemonizes and monitors all other organs.",
+    cloudNativeRole: "Control plane",
+    description: "Agent OS — supervises daemons, unified config, and health.",
     longDescription: "Agent Body acts as the master process supervisor for the Autonomic AI ecosystem. It daemonizes the other organs, monitors their process health, intercepts and unifies their logs, and handles automatic restarts if an organ crashes due to an out-of-memory error.",
     githubLink: "https://github.com/autonomic-ai-dev/agent-body",
     span: "col-span-1",
@@ -29,7 +31,8 @@ export const agents: AgentData[] = [
     name: "Agent Brain",
     icon: Brain,
     color: "from-indigo-500 to-purple-500",
-    description: "The core intelligence. Handles memory, context, and MCP routing.",
+    cloudNativeRole: "Memory store",
+    description: "Durable context, skills, and MCP routing.",
     longDescription: "Agent Brain is the memory and routing substrate. It solves context bloat by storing facts in a temporal SQLite knowledge graph and using filterable HNSW vector search to instantly retrieve relevant rules and context. It also acts as the MCP router for standard AI editors.",
     githubLink: "https://github.com/autonomic-ai-dev/agent-brain",
     span: "md:col-span-2 md:row-span-2",
@@ -40,7 +43,8 @@ export const agents: AgentData[] = [
     name: "Agent Spine",
     icon: Layers,
     color: "from-slate-500 to-slate-400",
-    description: "Deterministic execution engine. Manages structured DAG workflows.",
+    cloudNativeRole: "Workflow engine",
+    description: "Deterministic DAG workflows with snapshots and HITL gates.",
     longDescription: "Agent Spine enforces structured execution through YAML-defined DAG pipelines. It ensures that every step in an AI workflow is deterministic, replayable, and isolated. It supports parallel fan-out, immutable state snapshots, and human-in-the-loop approval gates.",
     githubLink: "https://github.com/autonomic-ai-dev/agent-spine",
     span: "col-span-1",
@@ -51,7 +55,8 @@ export const agents: AgentData[] = [
     name: "Agent Heart",
     icon: Heart,
     color: "from-rose-500 to-pink-500",
-    description: "Maintenance daemon. Enforces token budgets and runs scheduled GC.",
+    cloudNativeRole: "Controller / GC",
+    description: "Token budgets, scheduled GC, and background maintenance.",
     longDescription: "Agent Heart acts as the financial governor and garbage collector of the ecosystem. It strictly enforces token budgets to prevent runaway LLM costs, and runs periodic CRON routines to deduplicate facts, vacuum vector indexes, and clear expired memories.",
     githubLink: "https://github.com/autonomic-ai-dev/agent-heart",
     span: "col-span-1",
@@ -62,7 +67,8 @@ export const agents: AgentData[] = [
     name: "Agent Nerves",
     icon: Radio,
     color: "from-emerald-500 to-teal-500",
-    description: "Async event bus. Coordinates organs via NATS JetStream.",
+    cloudNativeRole: "Service mesh",
+    description: "NATS JetStream agent mesh and event routing.",
     longDescription: "Agent Nerves acts as the high-speed, decentralized communication bus between all organs. It manages a local NATS JetStream server, allowing organs to publish state changes and subscribe to relevant events with guaranteed at-least-once durable delivery.",
     githubLink: "https://github.com/autonomic-ai-dev/agent-nerves",
     span: "col-span-1",
@@ -73,7 +79,8 @@ export const agents: AgentData[] = [
     name: "Agent Muscle",
     icon: Cpu,
     color: "from-orange-500 to-amber-500",
-    description: "Execution actuator. Runs commands via JSON contracts and fine-tunes models.",
+    cloudNativeRole: "Execution runtime",
+    description: "Sandboxed commands, JSON contracts, and training jobs.",
     longDescription: "Agent Muscle handles physical actuation on the host machine. It executes shell commands enforcing strict JSON output schemas to prevent parsing errors. It also orchestrates local LoRA fine-tuning jobs using pluggable backends like Apple MLX or Candle.",
     githubLink: "https://github.com/autonomic-ai-dev/agent-muscle",
     span: "col-span-1",
@@ -84,7 +91,8 @@ export const agents: AgentData[] = [
     name: "Agent Immune",
     icon: Shield,
     color: "from-blue-500 to-cyan-500",
-    description: "Security layer. Scans dependencies and runs sandboxed execution.",
+    cloudNativeRole: "Admission / policy",
+    description: "OSV scanning, sandbox execution, and memory audits.",
     longDescription: "Agent Immune protects the ecosystem with three-layer defense: it queries OSV databases for dependency vulnerabilities, enforces seccomp or Firecracker isolation for untrusted LLM code, and audits post-execution memory logs for OOM violations.",
     githubLink: "https://github.com/autonomic-ai-dev/agent-immune",
     span: "col-span-1",
@@ -95,7 +103,8 @@ export const agents: AgentData[] = [
     name: "Agent Eyes",
     icon: Eye,
     color: "from-yellow-400 to-orange-400",
-    description: "Visual observability. Provides DOM indexing and local VLM inference.",
+    cloudNativeRole: "Observability",
+    description: "DOM index, capture, pixel diff, and visual QA.",
     longDescription: "Agent Eyes allows the ecosystem to reason about visual state. It parses UI layouts into DOM indexes, captures Playwright screenshots, computes structural pixel diffs via SSIM, and runs native, local LLaVA vision-language inference to identify visual regressions.",
     githubLink: "https://github.com/autonomic-ai-dev/agent-eyes",
     span: "col-span-1",
@@ -106,7 +115,8 @@ export const agents: AgentData[] = [
     name: "Agent Mouth",
     icon: MessageSquare,
     color: "from-teal-400 to-emerald-400",
-    description: "Communication daemon. Handles webhooks and log summarization.",
+    cloudNativeRole: "Ingress / gateway",
+    description: "AST validation, webhooks, approvals, and log summarization.",
     longDescription: "Agent Mouth routes outbound signals. When an organ triggers a critical event (like a security alert or deployment success), Mouth translates it into semantic payloads for configured webhooks (Slack, Discord). It can also synthesize dense JSON logs into human-readable summaries.",
     githubLink: "https://github.com/autonomic-ai-dev/agent-mouth",
     span: "col-span-1",

src/routes/+layout.svelte

@@ -15,7 +15,7 @@
 
 <svelte:head>
   <title>Autonomic AI Dev | Autonomous Agent Ecosystem</title>
-  <meta name="description" content="Autonomic AI Dev builds advanced, self-governing autonomous agents. Explore Brain, Nerves, Heart, Muscle, and the CI dashboard ecosystem." />
+  <meta name="description" content="Autonomic AI — cloud-native infrastructure for AI agents. Control plane, memory store, workflow engine, and agent mesh." />
 </svelte:head>
 
 <div class="min-h-screen relative overflow-hidden bg-slate-50 dark:bg-slate-950 font-sans selection:bg-brand-500/30">

src/routes/+page.svelte

@@ -17,12 +17,12 @@
   </div>
 
   <h1 class="text-5xl md:text-7xl font-extrabold tracking-tight text-white mb-6 leading-tight">
-    Autonomous systems that <br/>
-    <span class="text-gradient">build, heal, and evolve.</span>
+    Kubernetes for <br/>
+    <span class="text-gradient">AI agents.</span>
   </h1>
 
   <p class="text-lg md:text-xl text-slate-400 max-w-2xl mb-10 leading-relaxed">
-    Autonomic AI Dev is a self-governing ecosystem of specialized AI agents working together to orchestrate massive scale software engineering.
+    Cloud-native AI infrastructure — orchestration, durable memory, mesh networking, and policy for autonomous software engineering. Built in Rust, runs locally.
   </p>
 
   <div class="flex flex-col sm:flex-row items-center gap-4">
@@ -61,6 +61,7 @@
             <Icon class="text-white" size={24} />
           </div>
           <h3 class="text-2xl font-bold text-white mb-2">{agent.name}</h3>
+          <p class="text-xs font-semibold uppercase tracking-wider text-brand-400 mb-2">{agent.cloudNativeRole}</p>
           <p class="text-slate-400 group-hover:text-slate-300 transition-colors leading-relaxed">{agent.description}</p>
         </div>
 

src/routes/agents/[slug]/+page.svelte

@@ -27,6 +27,7 @@
 
       <div>
         <h1 class="text-4xl md:text-5xl font-extrabold text-slate-900 dark:text-white mb-3 tracking-tight">{agent.name}</h1>
+        <p class="text-sm font-semibold uppercase tracking-wider text-brand-500 dark:text-brand-400 mb-2">{agent.cloudNativeRole}</p>
         <p class="text-xl text-slate-600 dark:text-slate-300 font-medium">{agent.description}</p>
       </div>
 

src/routes/docs/+layout.svelte

@@ -35,6 +35,7 @@
     <div class="mb-8">
       <h3 class="text-sm font-bold text-slate-900 dark:text-white uppercase tracking-wider mb-3">Core Concepts</h3>
       <ul class="flex flex-col gap-2">
+        <li><a href="/docs/platform" onclick={() => isMobileDocsMenuOpen = false} class="{linkClass('/docs/platform')} text-sm flex items-center gap-2"><Layers size={16} /> Cloud-Native Platform</a></li>
         <li><a href="/docs/architecture" onclick={() => isMobileDocsMenuOpen = false} class="{linkClass('/docs/architecture')} text-sm flex items-center gap-2"><Layers size={16} /> Architecture</a></li>
         <li><a href="/docs/supervision" onclick={() => isMobileDocsMenuOpen = false} class="{linkClass('/docs/supervision')} text-sm flex items-center gap-2"><Server size={16} /> Supervision (Body)</a></li>
         <li><a href="/docs/memory" onclick={() => isMobileDocsMenuOpen = false} class="{linkClass('/docs/memory')} text-sm flex items-center gap-2"><Brain size={16} /> Memory (Brain)</a></li>

src/routes/docs/+page.md

@@ -8,7 +8,9 @@ Welcome to the **Autonomic AI** documentation.
 
 ## What is Autonomic AI?
 
-Autonomic AI is an open-source Rust-based multi-agent system for self-healing autonomous code maintenance and software lifecycle management. It arranges 9 specialized agent processes ("organs") into a coherent, event-driven system that can plan, execute, verify, and repair code without human supervision.
+Autonomic AI is **cloud-native AI infrastructure** for autonomous software engineering — a local control plane with specialized daemons (codename: *organs*) that plan, execute, verify, and repair code over an event-driven mesh.
+
+See [Cloud-Native Platform](/docs/platform) for the Kubernetes analogy and component mapping.
 
 The project lives at [github.com/autonomic-ai-dev](https://github.com/autonomic-ai-dev).
 

src/routes/docs/communication/+page.md

@@ -4,7 +4,7 @@ title: Notifications & Logging
 
 # Communication (agent-mouth)
 
-The final piece of the biological architecture is the ability to speak. `agent-mouth` is the dedicated communication and notification daemon for the Autonomic ecosystem.
+`agent-mouth` is the **I/O gateway** — AST command validation, webhooks, approvals, and log summarization for the Autonomic stack.
 
 ## Webhooks & Integrations
 

src/routes/docs/event-bus/+page.md

@@ -4,7 +4,7 @@ title: Event Bus
 
 # The Event Bus (agent-nerves)
 
-The Autonomic ecosystem is fundamentally decentralized. The 9 core "organs" do not communicate with each other via direct HTTP calls or shared memory. Instead, they operate as entirely independent processes that communicate asynchronously via a central event bus.
+The Autonomic stack uses an **agent mesh** — independent daemons communicate asynchronously via NATS JetStream, not direct HTTP coupling between every pair.
 
 This event bus is managed by the `agent-nerves` daemon, which acts as a bridge to the underlying [NATS server](https://nats.io/) infrastructure.
 

src/routes/docs/execution/+page.md

@@ -4,7 +4,7 @@ title: Execution & Actuation
 
 # Execution (agent-muscle)
 
-In an AI system, "thinking" is safe, but "acting" is dangerous. The `agent-muscle` organ is the dedicated execution actuator for the Autonomic ecosystem. It is responsible for translating the decisions made by the `agent-brain` and `agent-spine` into real-world state changes on the host machine.
+`agent-muscle` is the **execution runtime** (kubelet / CRI analog) — structured command execution with JSON contracts on the host.
 
 ## Structured Contracts
 

src/routes/docs/maintenance/+page.md

@@ -4,7 +4,7 @@ title: Maintenance & Budgeting
 
 # Maintenance (agent-heart)
 
-A self-healing system requires a heartbeat. The `agent-heart` organ is a specialized background daemon responsible for the long-term health, garbage collection, and financial safety of the ecosystem.
+`agent-heart` is the **background controller** — scheduled GC, token budget gates, and long-term memory hygiene for the stack.
 
 ## Garbage Collection (GC)
 

src/routes/docs/platform/+page.md

@@ -0,0 +1,32 @@
+---
+title: Cloud-Native Platform
+---
+
+# Kubernetes for AI Agents
+
+Autonomic AI is **cloud-native AI infrastructure** — a local control plane that gives agentic workloads what containers got in 2014: orchestration, durable state, memory, mesh networking, policy, and observability.
+
+## Platform mapping
+
+| K8s primitive | Component | Binary | Role |
+|---------------|-----------|--------|------|
+| Control plane | body | `autonomic` | Agent OS — lifecycle, config, supervisor |
+| PersistentVolumes + ConfigMaps | brain | `agent-brain` | Memory store |
+| Jobs / Argo / StatefulSets | spine | `agent-spine` | Workflow engine |
+| Service mesh | nerves | `agent-nerves` | Agent mesh (NATS) |
+| Admission / policy | immune | `agent-immune` | Zero-trust policy |
+| Controller / GC | heart | `agent-heart` | Background controller |
+| kubelet / CRI | muscle | `agent-muscle` | Execution runtime |
+| Observability | eyes | `agent-eyes` | State extraction |
+| Ingress / gateway | mouth | `agent-mouth` | I/O gateway |
+| Conformance (Sonobuoy) | benchmarks | `agent-benchmarks` | Ecosystem conformance |
+
+> **Codenames:** *Organ* names are stable internal identifiers in repos and `~/.autonomic/config.toml`. Enterprise docs lead with cloud-native roles above.
+
+Full reference: [agent-body cloud-native-platform.md](https://github.com/autonomic-ai-dev/agent-body/blob/master/docs/cloud-native-platform.md)
+
+## What we are not
+
+- Not a replacement for cluster Kubernetes — we orchestrate **agents**, not pods
+- Not a hosted SaaS — data lives under `~/.autonomic/` on your machine
+- Not a monolithic Python agent loop — independent Rust daemons with strict interfaces

src/routes/docs/supervision/+page.md

@@ -4,7 +4,7 @@ title: Supervision & Daemonization
 
 # Supervision (agent-body)
 
-The entire Autonomic ecosystem runs on a foundation of reliability and daemonization provided by `agent-body`. Rather than manually starting individual servers in separate terminal tabs, `agent-body` acts as the master supervisor process.
+The Autonomic stack runs on a **control plane** (`agent-body`) that supervises daemons with health probes and ordered boot — like `kubectl` plus a local supervisor.
 
 ## The Process Tree
 

src/routes/docs/visual-qa/+page.md

@@ -4,7 +4,7 @@ title: Visual Observability
 
 # Visual QA (agent-eyes)
 
-Code changes don't just affect tests and logic; they affect the user interface. The `agent-eyes` organ provides the Autonomic ecosystem with visual observability, allowing the agent to "see" the UI changes it makes.
+`agent-eyes` provides **observability for visual state** — DOM indexing, capture, and regression checks for agent-driven UI work.
 
 ## DOM Indexing