Profile Model Managerv0.1.0

An AI operations team manages Hermes agents across multiple profiles — customer support, code review, and research — each with different cost and capability needs. Using Profile Model Manager, they batch-switch cheap profiles to budget models like glm-5-turbo while keeping reasoning-heavy profiles on premium models. This cuts inference costs by routing workloads to the right model tier.

When a provider suffers an outage or pricing change, teams need to migrate all Hermes profiles to a new provider quickly. Profile Model Manager enables previewing changes with a dry-run, bulk-switching provider, base_url, and model settings, then restarting Gateways. This reduces downtime from hours of manual config edits to minutes.

Developers building Hermes-based agents need to compare model behavior across providers during testing. Profile Model Manager lets them quickly swap main, auxiliary (vision, compression, web_extract, session_search), and delegation models per profile to benchmark quality and latency. This accelerates experimentation with different model combinations.

Teams using Hermes delegation features (sub-agents) often spend heavily on child-task model calls. Profile Model Manager allows setting a cheaper delegation model distinct from the main model, or clearing it to inherit the main model. This provides granular cost control for multi-agent workflows without sacrificing main-model quality.

When agents report errors like 'API key not found', 'switch not taking effect', or 'context too small', operators need to inspect model configs across profiles. Profile Model Manager provides an overview script and per-profile config views to diagnose which profile has a mismatched provider, missing API key, or insufficient context_length.

Distribute Profile Model Manager as an open-source Hermes skill to drive adoption, while offering a paid managed service that hosts multi-profile model orchestration, dashboards, and automated failover for enterprises. The hosted layer adds auditing, RBAC, and scheduled provider-switching policies.

Wrap Profile Model Manager into a broader cost-optimization platform that monitors token spend per profile and automatically recommends or applies cheaper model switches based on workload patterns. Integrates with billing APIs to show real-time savings from model rerouting decisions.

Package the skill as part of an enterprise AI governance and configuration-management suite that enforces approved model/provider policies across all Hermes profiles, with change approval workflows and compliance logs. Targets regulated industries needing audit trails for model usage.