v1.0.2

What's Changed

Fixed

• all skills — Closed indirect prompt injection attack surface (OWASP LLM01 / W011). Removed unconditional INDEX.md reads and inlined entity slug tables directly in each SKILL.md; added data-only preamble to all 395 reference files; added trust-boundary framing to the start skill issue fetch; added defensive data-source instructions to all three skill SKILL.md files; and added SHA-256 integrity manifest generation to catholic-advisor-setup after install. (#3)

Full Changelog: v1.0.1...v1.0.2