Activation
The Librarian Agent is currently available behind the feature flag EnableContentAutomationHub.
This feature flag is enabled by default in all Sandbox environments, allowing you to explore the Librarian Agent alongside your existing rules and gradually transition ingestion automation where it adds value.
If you want to enable the Librarian Agent in your Production environment earlier, you can submit a request via the Aprimo Helpdesk.
Starting in November, the Librarian Agent will be enabled in Production for all customers and will become the recommended approach for managing ingestion automation, replacing ingestion logic currently handled by rules.
When should I use the Librarian Agent instead of Rules?
For ingestion processing, the Librarian Agent is the recommended and preferred approach. Managing ingestion logic through rules can lead to duplicate processing, limited visibility, and increased complexity. By centralizing this logic in the Librarian Agent, you benefit from:
- more consistent and reliable outcomes
- clear visibility into how content is processed
- easier configuration and long term maintenance
Recommendation: Review existing rules triggered on file ingestion and consider moving them into the Librarian Agent where possible to avoid duplication and ensure a single, streamlined ingestion experience.
🚀 Migrate to the Librarian Agent
Move from managing rules to guiding an intelligent agent, step by step.
Step 1 — Identify ingestion rules
What to do
Review your existing rules and identify those triggered by:
- File added
- File version added
Why this matters
These rules typically contain ingestion logic that the Librarian Agent can handle more efficiently.
✅ Done when: You have a clear list of ingestion related rules
Step 2 — Map rules to agent capabilities
What to do
For each rule, determine how the Librarian Agent can achieve the same outcome, for example:
- Metadata enrichment → Predictive metadata
- Tagging → Smart tags
- Transcription → Audio or video capabilities
- Classification → Automatic content typing
Why this matters
You are shifting from rules to agent capabilities, not losing functionality.
✅ Done when: Each rule has a corresponding agent capability (or is no longer needed)
Step 3 — Enable and guide the agent
What to do
Enable the relevant capabilities and define conditions to guide how the agent behaves.
Examples:
- Content type = Video
- File type = MP4
- Only first version
Why this matters
Conditions ensure the agent applies actions only when relevant, keeping processing efficient and accurate.
✅ Done when: Capabilities are enabled and scoped correctly
Step 4 — Test and validate
What to do
Test your setup in a Sandbox environment:
- Upload sample files
- Review results in the file processing activity log
- Confirm expected outcomes
Why this matters
Validates that the Librarian Agent produces the same or better results than your rules.
⭐ Tip: Start with a small set of files for quick validation
✅ Done when: Results match expectations and no duplicate processing occurs
Step 5 — Transition from rules
What to do
Once validated:
- Disable rules triggered on file ingestion
- Or remove the File added trigger if still needed elsewhere
Why this matters
Prevents duplicate processing and ensures the Librarian Agent is the single source of ingestion automation.
⭐ Important: This is the most critical step to avoid conflicts
✅ Done when: No overlapping ingestion rules remain active
Step 6 — Enable downstream automation
What to do
Update rules that depend on ingestion results to use:
→ File ready for use trigger
Why this matters
Ensures workflows and automations only run after the Librarian Agent has completed processing.
✅ Done when: Downstream automations trigger after enrichment is complete
Step 7 — Monitor and optimize
What to do
After migration:
- Monitor processing using the activity log
- Refine conditions
- Enable additional capabilities where needed
Why this matters
The Librarian Agent improves over time as you fine-tune how it is guided.
💡 Tip: Start small and expand gradually
✅ Done when: Processing is stable, efficient, and aligned with your needs