MOAR AI def issues

[commjoen]

Context

• What should the challenge scenario be like?
  Agent instruction/context files: CLAUDE.md, AGENTS.md, .cursor/rules, .windsurfrules, etc. can contain real credentials added as persistent context and then accidentally committed. This generalizes the skill-file challenge to the wider ecosystem. We already have skills.md so now we need to add it to others.
• What should the participant learn from completing the challenge?
  be careful what you put in agents.md/rules/etc.
• For what category would the challenge be? (e.g. Docker, K8s, binary)
  AI

Did you encounter this in real life? Could you tell us more about the scenario?

yes