Agent Architecture Scorecard

Score your AI agent setup in 5 questions. Get personalized recommendations based on your architecture maturity.

1. Does your agent have a SOUL.md (or equivalent identity/voice file)?

Yes — fully defined with voice, rules, boundaries Partially — some personality but not structured No — my agent uses defaults

2. How many core config files does your agent have? (SOUL.md, AGENTS.md, USER.md, TOOLS.md, MEMORY.md, SECURITY.md, IDENTITY.md, HANDOFF.md, HEARTBEAT.md)

7-9 — full architecture 4-6 — decent setup 0-3 — barely started

3. Does your agent remember context between sessions?

Yes — persistent memory files that carry forward Sometimes — I paste context manually No — every session starts from scratch

4. Does your agent run on a schedule (cron jobs, automated tasks)?

Yes — multiple scheduled jobs running daily One or two — basic automation No — I trigger everything manually

5. Can your agent handle errors and recover without your help?

Yes — retry logic, fallbacks, error logging Partially — I sometimes have to step in No — errors break the workflow entirely