Skip to content
CXOps Cloud is a personal exploration project — not a product, not for sale, not accepting signups. Learn more.

Agent Configuration

CXOps Cloud uses specialized AI agents for different tasks across the platform. The Control console gives you visibility into how these agents operate and their current workload.

Each module has dedicated agents that handle background processing:

  • Analyze Agent — Extracts brand voice, tone, language, and terminology from uploaded documents using AI. Processes PDF, DOCX, and TXT files.
  • Index Agent — Creates vector embeddings of your brand guidelines for semantic matching. Powers brand consistency checks in Govern.
  • Refresh Agent — Polls RSS feeds, calls the DataForSEO API, and generates ambient signals. Runs on your configured schedule (manual, daily, weekly, or real-time).
  • Optimize Agent — The core content generation agent. Scrapes source URLs, runs SEO/GEO analysis, and generates optimized content sections using AI. Respects your brand voice and persona settings.
  • Re-optimize Agent — Scheduled agent that re-runs optimization on projects with auto-optimize enabled. Runs nightly.
  • Check Agent — Runs compliance evaluations. Performs semantic similarity scoring for brand consistency, term matching for terminology rules, and tone detection for voice compliance.
  • Sync Agent — Handles GA4 OAuth token refresh and metric import from Google Analytics.
  • Aggregate Agent — Computes metric aggregations and trend data for the performance dashboard.
  • Content Agent — Generates channel-specific content adaptations. Takes approved drafts and rewrites them for each configured channel, respecting constraints (character limits, required elements, tone overrides).

All agents use the same underlying AI model for consistency across the platform. Content generation, compliance checking, and channel adaptation all draw from the same model, ensuring uniform quality.

In the Control console, you can:

  1. View queue stats — see how many jobs each agent type has across all statuses
  2. Filter by product — focus on a specific module’s agents
  3. Check fail reasons — when an agent encounters an error, the specific reason is logged
  4. Retry failed jobs — requeue jobs that failed due to transient errors
  5. Track processing time — see created, started, and finished timestamps for each job