Add Redis Iris capabilities to How agents work

[mich-elle-luna]

• Add a "Managed agent context with Redis Iris" section that maps agent needs to Agent Memory, LangCache, Context Retriever, and Data Integration, with preview status and a link to the request flow.
• Mention LangCache, Context Retriever access tags, and Agent Memory sensitive-data exclusions in the performance, security, and cost sections.
• Link the Iris concepts page and interactive demos from Next steps.
• Rebuild "Types of agents you can build" as one row-ordered grid with the agent builder types first and two bullets per box.

---

Note

Low Risk
Documentation-only updates to agent concepts; no runtime or security code changes.

Overview
How agents work now positions Redis Iris as the managed alternative to hand-built context layers, with a new comparison table mapping agent needs (memory, semantic caching, governed data access, live sync) to DIY Redis patterns versus Agent Memory, LangCache, Context Retriever, and Data Integration, plus preview status and a link to the Iris request-flow doc.

The Types of agents you can build grid is flattened into a single row-ordered layout: agent-builder types lead, each card is trimmed to two bullets, styling is unified, and Content generation agents is moved to the end of the list.

Production guidance and Next steps gain Iris cross-links—LangCache in performance and cost, Context Retriever access tags and Agent Memory sensitive-data exclusions in security, and Iris concepts plus browser interactive demos under Learn more.

^{Reviewed by Cursor Bugbot for commit efc443c. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

🧠 Redis Memory

Found 9 related items from repository history (4 new this commit):

• PR — Enhance guide with Redis Agent Memory information

  Added note about Redis Agent Memory and Redis Iris for managed service options.

• PR — Add Context Engine (Redis Iris) agent type to the AI agent builder

  Adds a fourth agent type, "Context Engine Agent", whose memory is fully managed by the Redis Iris Context Engine instead of raw Redis. Each

• Commit — Add Context Engine (Redis Iris) agent type to the AI agent builder

  Add Context Engine (Redis Iris) agent type to the AI agent builder

• Commit — Add agent capabilities and Data Integration to When to choose Redis

  Add agent capabilities and Data Integration to When to choose Redis

• Commit — Enhance guide with Redis Agent Memory information

  Enhance guide with Redis Agent Memory information

• Commit — Add Redis Iris capabilities to How agents work 🆕

  Add Redis Iris capabilities to How agents work

• PR — Restyle Iris/AI context-engine tiles and update self-managed messaging 🆕

  • Add a tile-card shortcode matching the integrate page's card style (colored swatch, uppercase mono label, description, "Learn more" link)

• PR — Iris concepts 🆕

  Add four new concepts documents for Iris, RAM, Langcache and CR.

• PR — Replace Iris tile swatches with icons and link demos and agent builder 🆕

  • Add an optional icon parameter to the tile-card shortcode; tiles without it keep the colored swatch.

Memory updated at efc443c

[github-actions]

Staging links:
https://redis.io/docs/staging/update-agent-doc/develop/ai/agent-builder/agent-concepts/

[andy-stark-redis]

Minor (ignorable) suggestion but otherwise LGTM.