Feature the RMP architecture diagram and refresh the README.

Add the Hyper-AI-Lab architecture hero image, document the galaxy web stack and canary soft-fail behavior, and replace the old flow chart with an up-to-date mermaid control-plane diagram.

Author: Hyper-AI-Lab

README.md

@@ -1,12 +1,14 @@
 # Aura
 
+![Aura: The Reliability & Memory Plane for Production Slack Agents](docs/assets/AI_Agent_Reliability_Architecture.jpg)
+
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
 [![Python 3.12+](https://img.shields.io/badge/python-3.12+-blue.svg)](requirements.txt)
 [![OpenClaw](https://img.shields.io/badge/runtime-OpenClaw-0ea5e9.svg)](https://github.com/openclaw/openclaw)
 [![Org](https://img.shields.io/badge/org-Hyper--AI--Lab-111827.svg)](https://github.com/Hyper-AI-Lab)
 
 **Aura is a reliability & memory control plane for a production Slack agent.**  
-It sits beside [OpenClaw](https://github.com/openclaw/openclaw) and turns every user turn into a durable Temporal workflow: intake routing, process memory, evidence gates, idempotent Slack delivery, and multi-key LLM orchestration.
+It sits beside [OpenClaw](https://github.com/openclaw/openclaw) and turns every user turn into a durable Temporal workflow: intake routing, process memory, evidence gates, idempotent Slack delivery, multi-key LLM orchestration, and a routed galaxy web-research stack.
 
 Built by [Hyper-AI-Lab](https://github.com/Hyper-AI-Lab) · Homepage: [hyperailab.com](https://hyperailab.com/)
 
@@ -20,6 +22,7 @@ Chat-agent stacks are great at tools and models — and terrible at **ops truth*
 - Sessions forget process context across turns  
 - Rate limits stall the whole agent with no fair key rotation  
 - “Done” is whatever the model claimed, not what evidence allows  
+- Canary restarts can kill mid-flight user work if remediation is too aggressive  
 
 Aura (the **Reliability & Memory Plane**, RMP) is the sidecar that owns those guarantees while OpenClaw stays the execution engine.
 
@@ -30,56 +33,98 @@ Aura (the **Reliability & Memory Plane**, RMP) is the sidecar that owns those gu
 | Durable task intake | 3-layer funnel (fast path → vector gate → LLM classify) with `off` / `shadow` / `enforce` modes |
 | Workflow control plane | Temporal `GenericTask` / `CatalogTask` workflows, child steps, reconciler + janitor |
 | Process memory | Process-scoped recall + promotion; Qdrant vectors (`nv-embed-v1`) |
-| Slack ownership | OpenClaw plugin routes DMs to RMP; RMP posts the final reply (no double-send) |
-| LLM orchestration | Balanced NVIDIA key rotation, concurrency caps, cooldowns, usage ledger |
-| Production gates | Readiness API, hourly canaries, patch verify, backup/restore runbooks |
+| Slack ownership | OpenClaw plugin routes DMs to RMP; RMP posts the final reply (no double-send / no native fallback) |
+| LLM orchestration | Balanced NVIDIA key rotation, concurrency caps, fast idle rotate (~5s), usage ledger |
+| Galaxy web stack | Brave + LangSearch search; Jina Reader; Crawl4AI / Scrapling / Crawlee / ScrapeGraphAI; OpenClaw `browser` + browser-use + Obscura CDP |
+| Web capability routing | Intake analyzer picks `search` / `fetch` / `crawl` / `extract` / `interact` and injects a tool brief |
+| Production gates | Readiness API, hourly canaries with **soft-fail deferral** (no worker restart while user tasks run), orphan-reply Slack recovery |
 
 ## Architecture
 
+Hero diagram (full system narrative):
+
+<p align="center">
+  <img src="docs/assets/AI_Agent_Reliability_Architecture.jpg" alt="Aura RMP architecture" width="100%" />
+</p>
+
+Full-resolution PNG: [`docs/assets/AI_Agent_Reliability_Architecture.png`](docs/assets/AI_Agent_Reliability_Architecture.png)
+### Control-plane flow (current)
+
 ```mermaid
-flowchart LR
-  User["Slack user"] --> OC["OpenClaw gateway"]
-  OC --> Plugin["rmp_adapter plugin"]
-  Plugin -->|"POST /tasks"| API["RMP FastAPI"]
-  API --> Temporal["Temporal workflows"]
-  Temporal --> Worker["rmp-worker"]
-  Worker -->|"hooks/agent"| OC
+flowchart TD
+  User["Slack_user_DM"] --> OC["OpenClaw_gateway"]
+  OC --> Plugin["rmp_adapter_claim"]
+  Plugin -->|"POST_/tasks"| API["RMP_FastAPI"]
+  API --> Intake["3_layer_intake"]
+  Intake --> WebCap["WebCapabilityAnalyzer"]
+  Intake --> Mode{"execution_mode"}
+  Mode -->|conversational_or_structured| Generic["GenericTaskWorkflow"]
+  Mode -->|interact_gated| Catalog["CatalogTask_browser_automation"]
+  WebCap -.->|preferred_tools_brief| Generic
+  WebCap -.->|preferred_tools_brief| Catalog
+  Generic --> Worker["rmp_worker"]
+  Catalog --> Worker
+  Worker -->|"hooks/agent_rmp_task"| OC2["OpenClaw_execution"]
+  OC2 --> Tools["Tools"]
+  Tools --> Native["web_search_web_fetch_browser"]
+  Tools --> AuraWeb["aura_web_plugin"]
+  AuraWeb --> LangSearch["LangSearch"]
+  AuraWeb --> Jina["Jina_Reader"]
+  AuraWeb --> Stack["web_stack_:8791"]
+  Stack --> Crawl4AI
+  Stack --> Scrapling
+  Stack --> Crawlee
+  Stack --> ScrapeGraph
+  Stack --> BrowserUse["browser_use"]
+  Stack --> Obscura["Obscura_CDP_:9222"]
   Worker --> PG["PostgreSQL"]
   Worker --> Qdrant["Qdrant"]
-  Worker -->|"notify_slack_user"| Slack["Slack DM"]
+  Worker -->|"notify_slack_user"| Slack["Slack_DM_idempotent"]
+  Canary["hourly_health_canary"] --> Sentinel["canary_sentinel"]
+  Sentinel -->|"soft_timeout_+_active_users"| Defer["defer_worker_restart"]
+  Sentinel -->|"hard_stale_or_code_sync"| Restart["restart_rmp_api_worker"]
+  Reconciler["reconciler"] -->|"orphan_OpenClaw_reply"| Slack
 ```
 
 | Layer | Role |
 | --- | --- |
-| **OpenClaw** | Slack socket, LLM/tools, isolated `rmp_task_*` sessions |
-| **RMP (`app/`)** | API, workflows, intake, memory, quota broker, evidence |
-| **Plugin (`plugins/rmp_adapter`)** | Intercepts Slack → creates RMP tasks; suppresses native double-posts |
+| **OpenClaw** | Slack socket, LLM/tools, isolated `rmp_task_*` sessions (execution only — not Slack delivery owner) |
+| **RMP (`app/`)** | API, workflows, intake, memory, quota broker, evidence, canary sentinel, reconciler |
+| **Plugin (`plugins/rmp_adapter`)** | Intercepts Slack → creates RMP tasks; suppresses native double-posts (fail closed) |
+| **Web (`plugins/aura_web`, `plugins/langsearch`, `web-stack/`)** | Multi-backend search/fetch/crawl/extract/browser tools + localhost FastAPI backends |
+
+**Binding rules:** every Slack DM goes through RMP; MiniMax M3 is the primary chat model; LLM idle silence fails fast (~5s) and rotates NVIDIA keys.
 
 Deep dive: [`ARCHITECTURE.md`](ARCHITECTURE.md) · Runbooks: [`docs/runbooks/`](docs/runbooks/)
 
 ## Repository layout
 
 ```text
 aura/
-├── app/                  # FastAPI + Temporal + memory + intake
-├── plugins/rmp_adapter/  # OpenClaw plugin
-├── ops/                  # Canaries, backup, janitor, patch verify
-├── tests/                # Pytest suite
-├── docs/                 # Runbooks + planning history
-├── patch_openclaw.sh     # Re-apply dist patches after OpenClaw upgrades
-├── settings.example.json # Config template (no secrets)
-├── worker.py             # Temporal worker entrypoint
-└── ARCHITECTURE.md       # Full system design
+├── app/                     # FastAPI + Temporal + memory + intake + web routing
+├── plugins/
+│   ├── rmp_adapter/         # Slack claim → RMP tasks
+│   ├── aura_web/            # Galaxy web tools (Jina, Crawl4AI, …)
+│   └── langsearch/          # LangSearch web_search provider + API key holder
+├── web-stack/               # Local FastAPI backends + Obscura compose/systemd
+├── ops/                     # Canaries, backup, janitor, patch verify
+├── tests/                   # Pytest suite
+├── docs/                    # Runbooks, history, architecture assets
+├── patch_openclaw.sh        # Re-apply dist patches after OpenClaw upgrades
+├── settings.example.json    # Config template (no secrets)
+├── worker.py                # Temporal worker entrypoint
+└── ARCHITECTURE.md          # Full system design
 ```
 
 ## Quick start
 
 ### Prerequisites
 
-- Linux host (or VM) with Docker optional for Qdrant/observability  
+- Linux host (or VM) with Docker optional for Qdrant / Obscura / observability  
 - Python 3.12+, Node.js ≥ 22.23 (OpenClaw engines)  
 - PostgreSQL, Temporal, [OpenClaw](https://github.com/openclaw/openclaw) gateway  
 - NVIDIA NIM (or compatible) API keys for chat + embeddings  
+- Optional: Brave + [LangSearch](https://langsearch.com/) API keys; Obscura image `h4ckf0r0day/obscura`
 
 ### Setup
 
@@ -92,10 +137,13 @@ pip install -r requirements.txt
 cp settings.example.json settings.json
 # set api_key, production.slack_owner_user_id, vector/qdrant, task_registry.intake_mode
 
-# Link plugin into your OpenClaw plugins dir, then:
+# Link plugins into your OpenClaw plugins dir (rmp_adapter, aura_web, langsearch), then:
 bash patch_openclaw.sh
 bash ops/verify_openclaw_patch.sh
 
+# Optional galaxy web backends + Obscura CDP
+# systemctl enable --now aura-web-backends aura-obscura
+
 # Start API + worker (systemd units or process manager of your choice)
 # then:
 make production-check
@@ -107,14 +155,18 @@ make production-check
 make production-check   # health + OpenClaw patch verify + intake canaries
 make canary             # manual E2E canary task
 pytest -q               # unit/integration tests
+curl -s http://127.0.0.1:8791/health   # web-stack backends (if enabled)
 ```
 
 ## Configuration notes
 
 - **Never commit** `settings.json`, `.env`, auth profiles, or `data/`.  
 - Example config: [`settings.example.json`](settings.example.json).  
-- After every `npm install -g openclaw`, re-run `patch_openclaw.sh` (hook persistence, Slack suppress, allowUnsafe passthrough for RMP sessions).  
-- Model stack (typical): MiniMax M3 primary → DeepSeek V4 Flash → GLM-5.2; intake/subagents on DeepSeek Flash.
+- LangSearch / Jina keys live in OpenClaw `plugins.entries.*` (not this repo).  
+- Obscura remote mode: `OBSCURA_CDP_URL=http://127.0.0.1:9222` (Hermes-compatible).  
+- After every `npm install -g openclaw`, re-run `patch_openclaw.sh` (hook persistence, Slack suppress, allowUnsafe passthrough, ~5s LLM idle).  
+- Model stack (typical): MiniMax M3 primary → DeepSeek V4 Flash → GLM-5.2; intake/subagents on DeepSeek Flash.  
+- Health canary **soft** failures (`timeout`/`failed`) defer worker restart while user tasks are active; reconciler can recover finished OpenClaw replies to Slack if delivery was interrupted.
 
 ## Status