--- name: docs-status-page-reader description: Use when the platform or a specific Louis component (API, chat, OCR, drafting engine, citation lookup) appears degraded or unavailable, and the system or a user needs to surface the current operational status. Drives in-product status banners, support triage, and incident communication by reading the live status page and reporting per-component health. license: MIT metadata: id: docs.status-page-reader category: docs jurisdictions: [__multi__] priority: P2 intent: [status, incident, degraded service, uptime, system health] related: [docs-security-overview, docs-tenant-isolation-explainer, docs-sso-saml-setup] source: Louis — HAQQ Legal AI (github.com/sboghossian/mini-claude-for-legal) version: "1.0" --- # Status Page Reader — Louis Platform Health ## Purpose This skill surfaces real-time operational status for each Louis platform component. It is used in two contexts: 1. **In-product banners** — when the application detects latency or errors, it calls this skill to retrieve the authoritative status and display a user-friendly notice rather than a generic error. 2. **Support triage** — when a user reports an issue, this skill is the first check: is this a platform incident or a user-specific problem? ## Status page URL `https://louis.haqq.ai/status` ## Platform components monitored | Component | Description | User impact when degraded | |---|---|---| | **API** | Core REST/GraphQL API layer | All features unavailable; login may fail | | **Chat** | AI chat interface and model inference | Conversations not loading or slow | | **OCR** | Document ingestion and optical character recognition | PDF/image uploads fail; extracted text unavailable | | **Drafting** | AI drafting engine | Draft generation fails or times out | | **Citations** | Legal citation lookup and validation | Citation insertion returns no results | ## Status levels | Level | Color | Meaning | |---|---|---| | **Operational** | Green | Component fully functional | | **Degraded Performance** | Yellow | Component functional but slower or partially limited | | **Partial Outage** | Orange | Some users or features affected | | **Major Outage** | Red | Component unavailable for all or most users | | **Under Maintenance** | Blue | Planned maintenance window | ## In-product behavior When the API or Chat component is degraded or worse: - Display a banner at the top of the application: > "Some features may be unavailable. Check [status.haqq.ai](https://louis.haqq.ai/status) for updates." - Do not suppress individual error messages — surface them alongside the banner. - Do not retry indefinitely; apply exponential back-off and surface a clear "Try again later" message after 3 failures. When OCR is degraded: - Show a specific message on the document upload dialog: > "Document processing is currently experiencing delays. Your upload will be queued and processed once the service recovers." When Drafting is degraded: - Disable the "Generate Draft" button and display: > "The drafting engine is temporarily unavailable. Please try again shortly." ## Support triage logic ``` User reports issue → Read status page → If any relevant component is not Operational: "We're aware of an issue affecting [component]. Our team is working on it. Follow https://louis.haqq.ai/status for updates." → Else: Collect user-specific diagnostics (browser, account, specific document, error message) Escalate to support ticket ``` ## Incident communication standards - **Detection to first update**: target < 15 minutes for Major Outage. - **Update frequency during incident**: every 30 minutes until resolved. - **Post-incident review**: published on the status page within 5 business days for Major Outage or incidents lasting > 2 hours. - **Customer notification**: Enterprise customers with SLA agreements receive direct email notification within 30 minutes of a Major Outage declaration. ## How to subscribe to status updates Users can subscribe to status page notifications at `https://louis.haqq.ai/status/subscribe`: - Email notifications for all incidents. - Webhook option for enterprise integrations (pipe alerts into Slack, PagerDuty, or your own monitoring). ## Related skills - [[docs-security-overview]] - [[docs-tenant-isolation-explainer]] - [[docs-sso-saml-setup]]