Skip to content

docs(wiki): add topic pages to the project wiki - #37

Merged
ksdisch merged 1 commit into
mainfrom
docs/wiki-pages
Jul 27, 2026
Merged

docs(wiki): add topic pages to the project wiki#37
ksdisch merged 1 commit into
mainfrom
docs/wiki-pages

Conversation

@ksdisch

@ksdisch ksdisch commented Jul 27, 2026

Copy link
Copy Markdown
Owner

What each page synthesizes and why it clears the no-restatement bar

Milestone-Feature-Map

Draws across 4 plan docs (m5-deploy-plan.md, m7-talents-plan.md, m8-strength-talents-plan.md, m9-themed-puzzles-plan.md) + BACKLOG Done entries + Decisions.md + AUDIT-2026-07-09.md. No single plan doc covers more than one milestone. This page answers: what did each milestone ship vs. deliberately cut, what decisions shaped those cuts (D2/D3/D5/D6/D8), and how the wire protocol grew across M5–M9. The protocol-growth section (optional fields, backward-compat design) is a pattern visible only when reading all four plan docs together.

Puzzle-and-Planet-Data-Model

Mined entirely from code (planet1.ts, registry.ts, protocol.ts, talents.ts, save.ts, portrait.ts, relay.ts). No docs file describes the full schemas or the relationships between PlanetConfig, PuzzleOverrides, ProgressState/PlanetTelemetry, and relayForward. The "0-is-no-sample" guard in both the clear-best and per-power bestMs accumulators, the relay allowlist's growth pattern, and the two-function architecture (tuningFor / strengthFor) are all code-only facts.

Deploy-And-Ops

Synthesizes DEPLOY.md (accurate walkthrough) + m5-deploy-plan.md (deliberate cuts: tsx at runtime, Cloudflare left out, relay-only container) + Decisions.md D6/D8 (Docker slim approach still unresolved; Phase 6 rides the deploy) + AUDIT Phase 6 findings (F-34 multi-machine HA, F-60 image size, F-61 Dockerfile hygiene) that DEPLOY.md does not mention. F-34 in particular (Fly default HA → rooms split across machines → joins fail) is the highest-priority pre-deploy finding not yet landed.

Rejected page

Autonomy-Loopdocs/AUTONOMY.md is already a complete, detailed standalone spec (the audit called it "impressively accurate"). Reading the test bridge source alongside it adds no independent synthesis; the doc references the code throughout and is accurate. A wiki page would restate rather than synthesize.

🤖 Generated with Claude Code

…net-Data-Model, Deploy-And-Ops

Each page synthesizes across multiple sources that no single docs file covers:
- Milestone-Feature-Map: draws across all four plan docs (m5/m7/m8/m9), BACKLOG Done
  entries, Decisions.md (D6/D8), and the audit to show what each milestone actually
  shipped vs. cut — and how the wire protocol grew across them.
- Puzzle-and-Planet-Data-Model: mines PlanetConfig, PuzzleOverrides, TALENTS, ProgressState/
  PlanetTelemetry, Portrait, and relayForward entirely from code; this schema exists in no
  docs file.
- Deploy-And-Ops: synthesizes DEPLOY.md (walkthrough), m5-deploy-plan.md (deliberate cuts),
  Decisions.md D6/D8, and Phase 6 audit findings (F-34 multi-machine HA and others) that
  DEPLOY.md does not mention.

Autonomy-Loop rejected: AUTONOMY.md is already a complete standalone spec;
code adds no independent synthesis. _index.md updated with all three new rows.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
@ksdisch
ksdisch merged commit b5a869d into main Jul 27, 2026
1 check passed
@ksdisch
ksdisch deleted the docs/wiki-pages branch July 27, 2026 02:08
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant