self.md radar — 2026-08-03
a public vault is only useful if it has a door, not a leak. Simon Späti has put 660 notes and 308,421 words online from a much larger Obsidian archive; the other 7,700 notes stayed private, while the published layer became a site, essays, and book chapters.
1. the public vault kept its door
sources:
what happened:
Späti updated a public account of his note system: 8,360 notes, about 3.07 million words, and 660 published notes containing 308,421 words. The published material comes from a private Obsidian vault via a #publish tag and a local deploy command; the public code uses Quartz and Hugo, then rsyncs to his own server. His own examples describe separate older notes converging later into a blog series or a book chapter, rather than a daily feed trying to impersonate thought.
the collision: The interesting boundary is not “second brain, but public.” It is editorial custody: a person can keep a body of work portable and searchable without turning every private fragment into content inventory. The tag is a small permission surface, but it changes the job from dumping a life online to making a public object someone else can actually use.
question left open: what would a personal archive need to preserve before an agent can help retrieve from it without quietly treating privacy as an indexing bug?
left on the table
- Aiden has an unusually explicit ledger for effects, approvals, evidence, and recovery, but it is a thin new repository claim in a week already crowded with personal-agent runtimes and verification machinery.
- EU age-verification discussion points toward a serious hardware-attestation boundary, but the cache supplied argument rather than a clean primary project record. That is not enough to publish the governance claim cleanly.
- Bor v0.8.0 is a concrete Linux policy-management release, but it does not yet move beyond ordinary device administration into a distinct human-control question.
Related self.md routes
- Personal AI OS tools — the control-plane map for personal agents, receipts, memory, and tools
- Best MCP servers — connect files, browsers, memory, search, and workflow tools without turning the stack into soup
- Agent memory systems — what agents should remember, what belongs in logs, and how to avoid memory sludge