Files
proof-of-work/start9/0.4
Keysat 974c3eb07d v1.1.0:2 — model-agnostic AI program generation (5 providers)
Five providers behind one streaming abstraction:
  - claude              (Anthropic)
  - openai              (api.openai.com)
  - openai-compatible   (any base URL — OpenRouter / LiteLLM /
                         vLLM / Together / your own gateway)
  - gemini              (Google)
  - ollama              (self-hosted; no key; LAN URL like
                         http://ollama.embassy:11434)

The "self-hosted Ollama on Start9" angle is the killer use case —
configure Settings → AI integration with the LAN URL of your Ollama
service and no API keys ever leave your network.

Architecture
  - lib/ai/types.ts              LLMProvider streaming interface
  - lib/ai/sse.ts                shared SSE + NDJSON line iterators
  - lib/ai/providers/*.ts        5 implementations + factory
  - lib/ai/programSchema.ts      Zod schema + JSON-schema-for-prompt +
                                  parseAIProgram with markdown-fence
                                  stripping and balanced-brace JSON
                                  extraction
  - lib/ai/apply.ts              materializes parsed AIProgram into
                                  Program tree (validates exerciseIds,
                                  rejects unresolved nulls, atomic
                                  transaction, sets aiGenerated=true)

Schema
  - UserPreferences gets aiProvider/aiModel/aiBaseUrl/aiApiKey
    (plaintext — same threat model as the rest of /data). Dead
    enableClaudeAI/claudeApiKey columns from v1.0.0:1-7 stay as
    no-op fields.
  - AIPromptTemplate (userId nullable; userId=NULL = built-in)
  - AIGeneration (raw response + parsed program + status +
    appliedProgramId + token counts)
  - All compat-ALTER'd in docker_entrypoint.sh on first boot.

API
  - POST   /api/ai/generate              SSE streaming: emits
                                           generation/text/usage/complete
                                           events; persists AIGeneration
                                           row up front so failures show
                                           in history too
  - POST   /api/ai/apply                 takes user-edited AIProgram,
                                           creates Program, marks
                                           generation as applied
  - GET    /api/ai/templates             built-ins + this user's own
  - POST   /api/ai/templates             create user-owned template
  - PATCH  /api/ai/templates/[id]        edit; built-ins admin-only
  - DELETE /api/ai/templates/[id]        delete; built-ins admin-only
  - GET    /api/ai/generations           list (paginated)
  - GET    /api/ai/generations/[id]      full row
  - DELETE /api/ai/generations/[id]      delete one (Program survives)
  - GET    /api/ai/config                returns aiKeyConfigured flag,
                                           never plaintext key
  - POST   /api/ai/config                update provider config
  - DELETE /api/admin/ai/generations     admin-only "clear all" with
                                           optional userId / olderThanDays

UI
  - Settings → AI integration            provider/model/URL/key form;
                                           plaintext key warning visible
  - /main/ai                             hub page with cards
  - /main/ai/generate                    template picker + textarea +
                                           live SSE stream + cancel +
                                           ProgramPreview with inline
                                           unknown-exercise resolver +
                                           apply button + redirect to
                                           the new Program
  - /main/ai/templates                   list + create + edit + delete;
                                           per-row "show prompt" expand;
                                           built-in delete warns about
                                           reconcile re-creation
  - /main/ai/history                     list + delete; status badges;
                                           link to applied Program
  - Nav: "AI" entry between Programs and Exercises (Sparkles icon)

Built-in templates
  - prisma/aiTemplates.seed.json: 5 starter templates (hypertrophy /
    strength / endurance / recovery / custom)
  - prisma/ensurePromptTemplates.cjs: per-boot reconcile,
    INSERT-or-UPDATE keyed on (userId IS NULL AND name=...);
    user-created templates never touched

Tests
  - tests/ai-programSchema.test.ts: extractJson + parseAIProgram
    edge cases (markdown fences, balanced braces, malformed JSON,
    Zod shape rejection, unresolved-exerciseId tolerance)
  - tests/ai-apply.test.ts: materializes valid AIProgram, rejects
    cross-user exerciseIds, rejects unresolved exercises, honors
    isActive flag
  - tests/routes-ai-templates.test.ts: built-in vs user permissions,
    cross-user template isolation, /api/ai/config plaintext-key safety,
    provider enum validation
  - 123 tests across 14 files, all passing.

No data migration. Existing /data is augmented with the new columns
+ tables only.
2026-05-10 15:35:35 -05:00
..
2026-05-09 13:40:58 -05:00

Proof of Work Logo

Proof of Work on StartOS 0.4 (migration package)

This directory packages Proof of Work (proof-of-work) for StartOS 0.4 beta. It is the cutover package that carries your 0.3.5 data across to a new x86_64 StartOS 0.4 host.

Upstream app lives at ../../proof-of-work/ in this repo. Legacy 0.3.5 package lives at ../0.3.5/ (kept intact; do not modify). Codex's WIP 0.4 scaffold lives at ../0.4/ (kept intact; superseded by this folder).

Goals

  • Keep the package id proof-of-work so StartOS recognizes it as the same service.
  • Keep the persistent data volume main mounted at /data.
  • Keep the SQLite database at /data/app.db.
  • Preserve every existing workout, set, exercise, and preference.
  • Ship x86_64 only for 0.4 beta (sideload target).

How data preservation works

  1. seed/data/app.db holds a one-time snapshot of /data from the live 0.3.5 host (currently 1 user, 348 workouts, 164 exercises, 5720 set logs).
  2. The Dockerfile bakes that snapshot into the image at /app/seed/data/.
  3. On first boot only/data/app.db missing AND /data/.seeded absent — docker_entrypoint.sh copies the seed into /data/ and writes a .seeded marker.
  4. On every subsequent boot, /data/ is the sole source of truth; the seed in the image is ignored.

See seed/README.md for the snapshot provenance and row counts.

Image runtime

Property Value
Base image node:20-alpine (multi-stage build)
App runtime Next.js standalone + Prisma + SQLite
Entrypoint /usr/local/bin/docker_entrypoint.sh (dumb-init wrapped)
Internal port 3000
Architectures x86_64 (beta)

Build and sideload

cd start9/0.4
npm ci
make clean
make x86      # outputs proof-of-work_x86_64.s9pk

Sideload via StartOS web UI or make install (requires ~/.startos/config.yaml). Step-by-step instructions are in DEPLOY_040.md.

What is unchanged from 0.3.5

  • Package id: proof-of-work
  • Volume id: main
  • Mount path: /data
  • DB path: /data/app.db
  • Health endpoint: /api/health
  • Compat ALTER TABLE block (idempotent; no-op on a current DB)

What is new in 0.4

  • TypeScript SDK manifest under startos/
  • ExVer version (0.1.0:18) replaces the 0.3.5 4-part 0.1.0.17
  • Seed-on-first-boot with a .seeded marker and stderr logging
  • alertUpdate warning users not to Uninstall to troubleshoot
  • Self-contained Dockerfile — no references to ../0.3.5/ or ../0.4/

Follow-up releases (planned, do not ship yet)

  • v0.1.0:19 — remove the COPY seed/data \u2026 line and the seed block from the entrypoint once the cutover is confirmed. Leaves seed/ on disk unreferenced.
  • v0.1.0:19 or v0.1.0:20 — add a StartOS Package Action change-admin-credentials that updates the User row in /data/app.db (bcryptjs, salt rounds 10) so you can rename/rotate the admin from the StartOS UI.