Daemon

Manages your public daemon profile — a living page of what you're working on, thinking about, reading, and building. It pulls data from your LifeOS system, runs it through a deterministic security filter so only publicly safe content survives, and deploys a static site. Workflows cover update, read, preview, and deploy.

You want a public presence that stays current without hand-editing a profile page every week, and without ever leaking private data. Your real context lives in LifeOS — goals, projects, ideas, identity — mixed with things that must never go public: contacts, finances, health, names, paths, credentials. Manually copying the safe parts is slow and one slip publishes something you can't take back. This skill aggregates the safe sources, blocks the sensitive ones at the code level, and gives you a preview-then-approve gate before anything ships.

The DaemonAggregator reads LifeOS sources and merges them into daemon-data.json; a deterministic SecurityFilter (pattern matching, not an LLM) strips names, paths, credentials, and internal refs; the deploy step builds a fully static site. Sensitive files are never opened by the aggregator at all.

Two-repo pattern: public framework + private content.

LifeOS SOURCES (private, read-only)
 TELOS/ (missions, goals, books, movies, wisdom)
 KNOWLEDGE/Ideas/ (title + thesis only)
 PROJECTS.md (public projects only)
 MEMORY/WORK/ (abstracted to topic themes)
 PRINCIPAL_IDENTITY.md (public bio data)
 │
 ├──[DaemonAggregator.ts]──→ Reads sources, merges with existing data
 │
 ├──[SecurityFilter.ts]──→ Deterministic code-level allowlist filter
 │ Strips names, paths, credentials, internal refs
 │ NOT an LLM filter — enforced by pattern matching
 │
 └──→ daemon-data.json → ~/Projects/daemon-dm/ (PRIVATE repo)
 │
 ├──[Tools/DeployGate.ts]──→ Deterministic pre-deploy gate (blocks on
 │ expired/ungated ephemera, street-address/ZIP/
 │ coordinate/home-area strings, unapproved
 │ real-time phrasing, credentials, private feed URLs)
 │
 └──[deploy.sh]──→ Copies JSON into framework → VitePress build → Cloudflare WORKER
 │
 ~/Projects/daemon/ (PUBLIC repo — forkable framework)
 │
 src/worker.ts (generic):
 • /daemon-data.json — served through the edge with expired
 ephemera STRIPPED at request time (status/now/offerings/
 requesting items with past `expires`; non-default location
 falls back to location_default)
 • /feed.json — live-activity items aggregated every 30 min
 (cron + lazy refresh) from PUBLIC sources only, configured
 in daemon-data.json `feeds` (rss | beehiiv | github | x);
 cached in KV FEED_KV
 • secrets (CF worker secrets, never in code): X_BEARER_TOKEN,
 BEEHIIV_API_KEY

 STRUCTURALLY EXCLUDED (never read):
 CONTACTS.md, FINANCES/, HEALTH/, TRAUMAS.md,
 KNOWLEDGE/People/, KNOWLEDGE/Companies/,

skills/Daemon/
├── SKILL.md (this file)
├── Tools/
│ ├── DaemonAggregator.ts (reads LifeOS sources → daemon-data.json)
│ └── SecurityFilter.ts (deterministic content sanitizer)
├── Workflows/
│ ├── UpdateDaemon.md (aggregate → preview → approve → deploy)
│ ├── ReadDaemon.md (read current daemon-data.json)
│ ├── PreviewDaemon.md (show diff without deploying)
│ └── DeployDaemon.md (bash deploy.sh from daemon-dm)
└── Docs/
 └── SecurityClassification.md (public/private data categories)

1. Private by default: All data is private until explicitly classified as public
2. Code-level enforcement: SecurityFilter.ts is deterministic pattern matching, NOT LLM judgment
3. Structural exclusion: Sensitive files (CONTACTS, FINANCES, HEALTH) are never opened by the aggregator
4. Defense in depth: Aggregator filter + SecurityFilter + pre-commit hook + manual approval
5. Fail closed: If uncertain, exclude the content

The DaemonAggregator reads from these LifeOS sources. Every source is read only if present (existsSync-guarded) — installs using the unified single-file TELOS.md layout won't have the per-file TELOS sources below, and the aggregator skips them cleanly (--sources shows per-source OK/MISSING status):

This skill is designed to be generic:

1. Fork the public Daemon repo
2. Create your own private data repo with daemon-data.json
3. Configure with your own blocked names/paths
4. The aggregator reads from standard LifeOS directory structure
5. Use deploy.sh to build and deploy to your own Cloudflare Pages

Example 1: Full update cycle

User: "update daemon"
→ Aggregates LifeOS data sources
→ Applies security filter (deterministic)
→ Shows preview diff to user
→ User approves
→ Writes daemon-data.json to daemon-dm → deploys static site

Example 2: Check what's current

User: "check daemon"
→ Reads daemon-data.json from daemon-dm
→ Shows section-by-section status

Example 3: Preview before committing

User: "preview daemon"
→ Runs aggregator in preview mode
→ Shows diff against current daemon-data.json
→ No writes, no deploys

• Two repos: Public framework (~/Projects/daemon/) and private content (~/Projects/daemon-dm/). The framework is forkable. The content is yours.
• deploy.sh runs DeployGate FIRST, then copies data into the framework at build time, then cleans up. A gate failure blocks the deploy — fix the data, never bypass the gate. Personal data never gets committed to the public repo.
• SecurityFilter is code, not prompts. If you need to add new blocked patterns, edit SecurityFilter.ts, not the workflow markdown.
• Ephemera needs expires. status, now (via now_meta), time-bound offerings/requesting items, and any non-default location all carry ISO expires fields. Three enforcement layers: DeployGate (blocks), the worker (strips at serve time), the dashboard (hides client-side). Refreshing a stale status = edit daemon-data.json with a new expires and run deploy.sh.
• Location doctrine: coarse by default, real-time by exception. City/region granularity only; street/ZIP/coordinate/home-area strings are gate-blocked. "Tonight/I'm at" phrasing requires an explicit realtime_approved: true on that item — reserved for events where the principal WANTS to be findable. No automatic GPS/calendar pipeline; location changes only on explicit command.
• The live feed is public-exhaust only. The worker polls blog RSS, Beehiiv (official API), YouTube RSS, GitHub public events, and the owner's own X posts — already-published content, zero privacy risk by construction. Never add a LifeOS-internal source to feeds.
• Beehiiv blocks RSS scrapers (403). The newsletter source uses the official Beehiiv API (type: "beehiiv" + BEEHIIV_API_KEY worker secret), not the /feed URL.
• Serving is edge-dynamic, page shell is static. /daemon-data.json and /feed.json are computed per-request by the worker (run_worker_first); the rest is static assets. Data changes still require deploy.sh; feed content refreshes itself.
• The upstream framework repo publishes a GENERATED template (merge-back doctrine). Its main branch is a clean generic template published from a scrubbed staging copy, never a working tree pushed directly: the template excludes analytics, licensed fonts, favicons, and all personal content, and its identity is data-driven (owner_name/owner_handle/fork_url in daemon-data.json). The publish workflow is maintainer-side and does not ship; on an installed system your daemon content lives in your own private repo and only the framework is forkable.