Move advisory board personas into .claude/ repo files

[jkbennitt]

Problem

The advisory board rule (.claude/rules/advisory-board.md) references a persona file stored in Caleb's local Claude auto-memory:

~/.claude/projects/-home-hubcaps-Projects-mem/memory/persona_advisory_board.md

This path is machine-specific and doesn't exist on anyone else's setup. The 19 advisory board lenses are invisible to other contributors.

Proposed Fix

Move the advisory board into the .claude/ system as either a rule or a skill:

Option A: Skill (recommended)

The advisory board is invoked on demand ("run this through the board"), which fits the skill model. Create .claude/skills/advisory-board.md containing the persona definitions and invocation instructions. Users trigger it explicitly.

Option B: Rule

Keep it as a rule in .claude/rules/advisory-board.md but inline the persona content directly instead of referencing an external file. Downside: rules are always loaded into context, and 19 personas is a lot of passive context.

Either way, the content needs to live in the repo, not in local Claude memory.

Action

@CalebisGross — grab the content from:

~/.claude/projects/-home-hubcaps-Projects-mem/memory/persona_advisory_board.md

and set it up as a .claude/skills/ or .claude/rules/ file in the repo.

[CalebisGross]

good catch @jkbennitt. went with Option A (skill).

• created .claude/skills/advisory-board.md with the full 19 personas
• slimmed down .claude/rules/advisory-board.md to a 4-line trigger that points to the skill instead of my local memory path

the rule stays so Claude still knows when to invoke the board proactively during planning/architecture decisions, but it's just a pointer now. all the actual content loads on demand via the skill, so it's not eating context in every conversation.

will get this on a branch and merged.