Capacity trend analysis and projections: disk growth rates, resource consumption patterns, and time-to-exhaustion estimates. AUTO-CONSULT when: disk usage exceeds 60%, memory usage is consistently high, user asks about server capacity or growth trends, or during periodic maintenance reviews.
AUTO-CONSULT when: you receive a warning that iteration limit is near, the same tool fails 3+ times in a row, you are stuck and cannot make progress, or the task requires code changes beyond log analysis.
Deployment procedures with pre-checks, verification, and rollback awareness. AUTO-CONSULT when: deploying services, running serverdeploy, updating Docker stack, or user asks to deploy/update.
Dev mode workflow: when to activate observe-only watch, how to exclude services during development, and how to detect active development. AUTO-CONSULT when: same service restarts repeatedly in short window, triage shows errors typical of rebuilds (connection refused, image not found), user mentions they are…
Systematic server diagnostics methodology. AUTO-CONSULT when: investigating service failures, performance issues, or unknown errors. Guides the diagnostic process from triage through root cause analysis.
Escalate problems to other agents in the network. AUTO-CONSULT when: you cannot resolve an issue after diagnosis, need human decision, encounter cross-domain problems, or need code-level investigation.
Severity classification and response procedures for server incidents. AUTO-CONSULT when: receiving alerts, webhook notifications, watch-mode detects issues, or multiple services are unhealthy simultaneously.
System maintenance procedures: disk cleanup, Docker pruning, log rotation, resource management. AUTO-CONSULT when: disk usage is high, user asks about cleanup, periodic maintenance needed, or servercleanup/serverprune results need interpretation.
Knowledge base: store and retrieve operational knowledge. AUTO-CONSULT when: resolving an incident (search for similar past incidents first), after fixing a non-trivial issue (save the solution), during capacity reviews (check historical trends), or when user asks about past events.
Post-incident analysis: build timeline, identify root cause, document resolution, and extract prevention measures. AUTO-CONSULT when: incident is resolved, user asks for incident summary or root cause analysis, or after any P0 incident resolution.
Remote server management and troubleshooting via SSH. AUTO-CONSULT when: remote site is down, remote service issues, nginx/PHP/database errors, SSL problems on remote server, or healthcheck reports remote issues.
Generate structured status reports. AUTO-CONSULT when: user asks for a status report, summary, or overview of the server state. Produces clear, formatted reports.
Security review methodology for server infrastructure. AUTO-CONSULT when: user asks for security check, audit, or hardening review, or when investigating suspicious activity.
Container dependency graph, restart order, and cascading failure handling. AUTO-CONSULT when: restarting multiple services, dealing with cascading failures, database unreachable, or planning a full stack restart.
★not rated 5 yesterdayA41 tokens
originalMIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: