3,748 tools found
Design an internal LLM platform with model promotion gates, controlled rollouts, serving infrastructure, telemetry, and governance.
Scale GPU inference on Kubernetes using request queues, model metrics, replica limits, and measured startup and spot-capacity tradeoffs.
Route LLM requests through a shared gateway with virtual keys, budgets, provider failover, caching, and streaming-aware proxy controls.
Plan QLoRA, distributed training, preference alignment, and adapter export with explicit hardware, data, and deployment prerequisites.
Reduce repeated LLM work with exact, semantic, and provider prompt caching while preserving access boundaries and response freshness.
Harden LLM applications with input checks, tool authorization, PII redaction, and explicit limits on injection and toxicity filters.
Harden Linux servers with reviewed SSH, account, firewall, kernel, permission, and audit settings while preserving recovery access.
Administer Linux hosts with distribution-aware package management, filesystem and process diagnostics, scheduled jobs, logs, and network checks.
Manage YYLO Ledger tasks, dependencies, and execution order through controller-routed CLI commands with explicit archive and merge safeguards.
Create Kubernetes configuration variants from reusable bases and overlays, with targeted patches, generated resources, and render-before-apply checks.
Harden Kubernetes workloads with restricted admission, security contexts, network policies, and least-privilege RBAC within an authorized scope.
Route bounded yes/no, choice, and score judgments to Jev while retaining writing, authorization, and escalated decisions in the main workflow.
