Skill: /stranger [argument]

Build an outside-observer portrait from vault evidence alone — a stranger's perspective with no insider charity, surfacing what a perceptive outsider would conclude.

Don't: protect Shane's self-image in the analysis. Don't draw on self-described identity — only what the vault evidence reveals.

Steps

1. Parse the argument/topic from Shane's request.
2. Follow Qwen Protocol with:

• task: "Vault access (bash only, no MCP tools): obsidian search query='TERM' limit=10, obsidian read file='Note Name' (no .md). Build an outside-observer portrait of Shane based on the vault, focusing on '[argument]' or broadly. What would a perceptive stranger conclude about how he thinks?"
• skill: "stranger"

1. Review Qwen's result, synthesize if needed, and present to Shane.

Fallback

If Qwen is unavailable:

1. If [argument] is provided, run obsidian search query='[argument]' limit=15 via bash; otherwise run broad topic searches via bash (e.g. obsidian search query='work' limit=10, obsidian search query='ideas' limit=10, obsidian search query='journal' limit=10)
2. Read 10–15 notes by running obsidian read file='[note name]' via bash for each, sampling across different topics, time periods, and note types
3. Adopt the perspective of a perceptive stranger reading these notes cold — no insider context, no charity
4. Build a portrait based purely on what the notes reveal:

• What does this person clearly care about most?
• What patterns of thinking are immediately visible?
• What does this person seem to be afraid of or avoiding?
• What would a stranger guess about their core beliefs, even unstated ones?
• What would seem strange, inconsistent, or surprising from the outside?

1. Be unflinching — a stranger has no reason to protect Shane's self-image
2. Present the portrait to Shane, with specific evidence from notes for each observation