docs: Add guide for configuring agent memory (knowledge & lore sections)

[madjin]

Context

Users struggle with agent memory/recall configuration. Need clearer guidance on "knowledge" and "lore" sections in agent JSON.

• Source: Discord discussion (2026-01-03)
• Priority: P2 (from issue-triage assessment)
• Impact: High (recurring onboarding pain; "memory doesn't work" complaints)

Evidence

• Discord: Users asking "How should knowledge and lore sections be added to JSON configurations?"
• Open question in daily facts: "How should knowledge and lore sections be added to JSON configurations to improve agent memory/recall?"
• Pattern: Configuration discoverability problems across multiple reports

Technical Details

• Component: Docs + Agent configuration schema
• Complexity: 2/5 (primarily documentation work)
• Required Expertise: Agent config schema, technical writing, examples

Proposed Implementation

1. Create docs page: "Configuring Agent Memory: Knowledge & Lore Guide"
2. Include working JSON examples for both knowledge and lore sections
3. Explain the difference between knowledge (factual) and lore (personality/backstory)
4. Add schema validation guidance (what's required vs optional)
5. Include troubleshooting section for common memory issues

Acceptance Criteria

• Documentation page exists with clear examples
• Both knowledge and lore sections covered
• At least one working agent template with memory configured
• FAQ/troubleshooting section included

---

Generated from elizaOS intelligence pipeline
Evidence date: 2026-01-06