Second Brain Hub

Problem

A second brain often becomes fragmented: capture, external saving, distillation, creation, inbox, review, and diagnosis are spread across habits or plugins, so users collect a lot but produce little. This skill treats Second Brain as the single entry point and handles requests through a fixed runtime contract, instead of exposing separate methodology skills.

How It Works and Boundaries

• Fixed loop: each round reads references/runtime-protocol.md, maintains a Hub Run Ledger, then routes to one workflow.
• Intent routing: language signals select the relevant workflow-*.md; if ambiguous, it asks one clarification question.
• Contract execution: pulls current preconditions and capabilities from route-contracts.json and capability-contracts.json; if tools fail, reads dependencies.json and chooses primary, fallback, or blocked.
• Write safety: reads references/writing-pipeline.md before writing, updating, moving, or deleting; deletion requires showing targets and per-item confirmation.
• Output control: complex scenarios use step-chain map cards, ordinary results use short completion cards, and contracts, ledgers, gates, and capability IDs are not exposed.

It fits Obsidian-based second-brain workflows, not plugin installation, themes, shortcuts, weather, or one-off factual lookup. Diagnosis only gives bottlenecks, evidence, and recommended scenarios without modifying the Vault.