@lifeaitools/rdc-skills 0.24.41 → 0.25.0

package/guides/infrastructure.md

@@ -1,179 +1,179 @@
-# Infrastructure Agent Guide — Base
-> Role-based context for infra/deployment/DevOps agents. Generic patterns across projects.
-
----
-
-## Rule 1: NEVER Work Around Broken Infrastructure
-
-When a service is unavailable, STOP and report BLOCKED. Do not:
-- Use curl when service APIs are down
-- Use workarounds or alternative approaches
-- Skip verification steps
-- Assume service will return to normal
-
-**Report immediately:**
-
-```
-BLOCKED: [service name] is not responding.
-
-Fix: [specific action for your project's infrastructure stack]
-
-I cannot proceed until this is resolved.
-```
-
----
-
-## Deployment Tools
-
-Check project overlay for:
-- Primary deployment platform (Coolify, Vercel, AWS, GCP, etc.)
-- MCP connectors or REST API access
-- Authentication method (API token, OAuth, etc.)
-- Credential location (clauth daemon, env vars, etc.)
-
----
-
-## Server Infrastructure
-
-The project specifies:
-- Server IP/hostname
-- Dashboard URL
-- Default region/availability zones
-- Server/environment IDs (UUIDs or identifiers)
-
----
-
-## DNS Rules
-
-The project specifies:
-- Wildcard DNS patterns vs individual records
-- Cloudflare or other DNS provider
-- Proxy rules (orange cloud, DNS-only, etc.)
-- SSL/TLS provisioning method
-
-Check overlay for **critical rules** — DNS misconfigurations break deployments.
-
----
-
-## Deployment Registry
-
-The project likely has a registry/database table tracking all deployments. Before ANY deploy, verify:
-```
-Lookup: <slug> or <domain>
-Returns: UUID, repo location, build command, build type, domain, status
-```
-
-**NEVER guess** UUIDs, domains, or build commands. Always look them up.
-
----
-
-## Watch Paths
-
-For monorepo deploys, the project specifies:
-- Watch path patterns per app type
-- Why watch paths matter (prevents unnecessary rebuilds)
-- How to set watch paths (usually via API or config)
-
-Without correct watch paths, every push triggers ALL apps to rebuild.
-
----
-
-## Build Types
-
-The project specifies supported build types:
-- Next.js monorepo
-- Vite / Node
-- Static HTML
-- Docker
-- etc.
-
-Check overlay for:
-- Build pack (nixpacks, docker, static, etc.)
-- Required environment variables
-- Node version constraints
-- Build command and install command
-
----
-
-## Environment Tiers
-
-The project specifies deployment environments:
-- **development** -- free to experiment
-- **staging** -- test before production
-- **production** -- live traffic, needs confirmation
-
-Check overlay for which tier each app is in and confirmation requirements.
-
----
-
-## Deploy Checklist
-
-Standard pattern:
-1. Check git status (divergence, unpushed commits)
-2. Confirm with user (especially for production)
-3. Push to trigger auto-deploy (if webhook configured)
-4. Verify deployment success (health check, status endpoint)
-5. Check cache headers (if behind CDN)
-
----
-
-## New App Deployment
-
-The project specifies:
-- DNS pattern (wildcard subdomains vs custom domains)
-- Coolify/deployment platform setup
-- GitHub repo connection
-- Watch paths configuration
-- Registry update requirement
-
----
-
-## Credential Safety
-
-- MCP connectors first (if available)
-- clauth daemon second (localhost:52437)
-- Never print keys to stdout
-- Never hardcode credentials
-- Never ask user for keys
-- If daemon is down: report BLOCKED
-
----
-
-## Git Workflow
-
-The project specifies:
-- Primary branch for features (develop, main, etc.)
-- Force-push rules (usually: NEVER force-push main)
-- Commit message format
-- Auto-commit patterns
-
----
-
-## Service Health Checks
-
-The project specifies how to health-check each service:
-- Ping endpoint
-- Status endpoint
-- Log location
-- Fallback if primary method fails
-
----
-
-## Troubleshooting Patterns
-
-The project specifies common issues and fixes:
-- 502/503 errors (check container logs, port mismatch)
-- Disk full (clean Docker cache)
-- SSL provisioning failures (check DNS config)
-- Build failures (check environment variables, Node version)
-
----
-
-## Specialist Context — Read Project Overlay
-
-Your task may require reading additional project-specific guides for:
-- Full deployment registry schema
-- Complete DNS rules (critical for subdomains)
-- Build type details
-- CI/CD pipeline configuration
-- Scaling and performance tuning
+# Infrastructure Agent Guide — Base
+> Role-based context for infra/deployment/DevOps agents. Generic patterns across projects.
+
+---
+
+## Rule 1: NEVER Work Around Broken Infrastructure
+
+When a service is unavailable, STOP and report BLOCKED. Do not:
+- Use curl when service APIs are down
+- Use workarounds or alternative approaches
+- Skip verification steps
+- Assume service will return to normal
+
+**Report immediately:**
+
+```
+BLOCKED: [service name] is not responding.
+
+Fix: [specific action for your project's infrastructure stack]
+
+I cannot proceed until this is resolved.
+```
+
+---
+
+## Deployment Tools
+
+Check project overlay for:
+- Primary deployment platform (Coolify, Vercel, AWS, GCP, etc.)
+- MCP connectors or REST API access
+- Authentication method (API token, OAuth, etc.)
+- Credential location (clauth daemon, env vars, etc.)
+
+---
+
+## Server Infrastructure
+
+The project specifies:
+- Server IP/hostname
+- Dashboard URL
+- Default region/availability zones
+- Server/environment IDs (UUIDs or identifiers)
+
+---
+
+## DNS Rules
+
+The project specifies:
+- Wildcard DNS patterns vs individual records
+- Cloudflare or other DNS provider
+- Proxy rules (orange cloud, DNS-only, etc.)
+- SSL/TLS provisioning method
+
+Check overlay for **critical rules** — DNS misconfigurations break deployments.
+
+---
+
+## Deployment Registry
+
+The project likely has a registry/database table tracking all deployments. Before ANY deploy, verify:
+```
+Lookup: <slug> or <domain>
+Returns: UUID, repo location, build command, build type, domain, status
+```
+
+**NEVER guess** UUIDs, domains, or build commands. Always look them up.
+
+---
+
+## Watch Paths
+
+For monorepo deploys, the project specifies:
+- Watch path patterns per app type
+- Why watch paths matter (prevents unnecessary rebuilds)
+- How to set watch paths (usually via API or config)
+
+Without correct watch paths, every push triggers ALL apps to rebuild.
+
+---
+
+## Build Types
+
+The project specifies supported build types:
+- Next.js monorepo
+- Vite / Node
+- Static HTML
+- Docker
+- etc.
+
+Check overlay for:
+- Build pack (nixpacks, docker, static, etc.)
+- Required environment variables
+- Node version constraints
+- Build command and install command
+
+---
+
+## Environment Tiers
+
+The project specifies deployment environments:
+- **development** -- free to experiment
+- **staging** -- test before production
+- **production** -- live traffic, needs confirmation
+
+Check overlay for which tier each app is in and confirmation requirements.
+
+---
+
+## Deploy Checklist
+
+Standard pattern:
+1. Check git status (divergence, unpushed commits)
+2. Confirm with user (especially for production)
+3. Push to trigger auto-deploy (if webhook configured)
+4. Verify deployment success (health check, status endpoint)
+5. Check cache headers (if behind CDN)
+
+---
+
+## New App Deployment
+
+The project specifies:
+- DNS pattern (wildcard subdomains vs custom domains)
+- Coolify/deployment platform setup
+- GitHub repo connection
+- Watch paths configuration
+- Registry update requirement
+
+---
+
+## Credential Safety
+
+- MCP connectors first (if available)
+- clauth daemon second (localhost:52437)
+- Never print keys to stdout
+- Never hardcode credentials
+- Never ask user for keys
+- If daemon is down: report BLOCKED
+
+---
+
+## Git Workflow
+
+The project specifies:
+- Primary branch for features (develop, main, etc.)
+- Force-push rules (usually: NEVER force-push main)
+- Commit message format
+- Auto-commit patterns
+
+---
+
+## Service Health Checks
+
+The project specifies how to health-check each service:
+- Ping endpoint
+- Status endpoint
+- Log location
+- Fallback if primary method fails
+
+---
+
+## Troubleshooting Patterns
+
+The project specifies common issues and fixes:
+- 502/503 errors (check container logs, port mismatch)
+- Disk full (clean Docker cache)
+- SSL provisioning failures (check DNS config)
+- Build failures (check environment variables, Node version)
+
+---
+
+## Specialist Context — Read Project Overlay
+
+Your task may require reading additional project-specific guides for:
+- Full deployment registry schema
+- Complete DNS rules (critical for subdomains)
+- Build type details
+- CI/CD pipeline configuration
+- Scaling and performance tuning