Add coding guidelines

.cursor/rules/coding-guidelines.mdc

@@ -0,0 +1,70 @@
+---
+description: Behavioral guidelines to reduce common LLM coding mistakes. Use when writing, reviewing, or refactoring code to avoid overcomplication, make surgical changes, surface assumptions, and define verifiable success criteria.
+alwaysApply: true
+---
+
+# Karpathy behavioral guidelines
+
+Behavioral guidelines to reduce common LLM coding mistakes. Merge with project-specific instructions as needed.
+
+**Tradeoff:** These guidelines bias toward caution over speed. For trivial tasks, use judgment.
+
+## 1. Think Before Coding
+
+**Don't assume. Don't hide confusion. Surface tradeoffs.**
+
+Before implementing:
+- State your assumptions explicitly. If uncertain, ask.
+- If multiple interpretations exist, present them - don't pick silently.
+- If a simpler approach exists, say so. Push back when warranted.
+- If something is unclear, stop. Name what's confusing. Ask.
+
+## 2. Simplicity First
+
+**Minimum code that solves the problem. Nothing speculative.**
+
+- No features beyond what was asked.
+- No abstractions for single-use code.
+- No "flexibility" or "configurability" that wasn't requested.
+- No error handling for impossible scenarios.
+- If you write 200 lines and it could be 50, rewrite it.
+
+Ask yourself: "Would a senior engineer say this is overcomplicated?" If yes, simplify.
+
+## 3. Surgical Changes
+
+**Touch only what you must. Clean up only your own mess.**
+
+When editing existing code:
+- Don't "improve" adjacent code, comments, or formatting.
+- Don't refactor things that aren't broken.
+- Match existing style, even if you'd do it differently.
+- If you notice unrelated dead code, mention it - don't delete it.
+
+When your changes create orphans:
+- Remove imports/variables/functions that YOUR changes made unused.
+- Don't remove pre-existing dead code unless asked.
+
+The test: Every changed line should trace directly to the user's request.
+
+## 4. Goal-Driven Execution
+
+**Define success criteria. Loop until verified.**
+
+Transform tasks into verifiable goals:
+- "Add validation" → "Write tests for invalid inputs, then make them pass"
+- "Fix the bug" → "Write a test that reproduces it, then make it pass"
+- "Refactor X" → "Ensure tests pass before and after"
+
+For multi-step tasks, state a brief plan:
+```
+1. [Step] → verify: [check]
+2. [Step] → verify: [check]
+3. [Step] → verify: [check]
+```
+
+Strong success criteria let you loop independently. Weak criteria ("make it work") require constant clarification.
+
+---
+
+**These guidelines are working if:** fewer unnecessary changes in diffs, fewer rewrites due to overcomplication, and clarifying questions come before implementation rather than after mistakes.

docs/relaunch-brief.md

@@ -1,40 +1,29 @@
 # CommunityRule relaunch
 
-A short high-level summary of what's being built, what it replaces, and
-how the cutover will work.
+A short high-level summary of what's being built, what it replaces, and how the cutover will work.
 
 ## What gets replaced
 
-The existing `CommunityRule` Cloudron app currently hosts three things in
-one container:
+The existing `CommunityRule` Cloudron app currently hosts three things in one container:
 
 - The static marketing site.
 - The Express + MySQL backend (rule drafting, publishing, OTP sign-in).
 - A Flask chatbot.
 
-All three retire together when the new app goes live. The chatbot is
-**not** being migrated or replaced in this stage.
+All three retire together when the new app goes live. The chatbot is **not** being migrated or replaced in this stage.
 
 ## What the new app is
 
-- A Next.js application with a Postgres database, packaged as a
-  Cloudron app (Docker image + `CloudronManifest.json`).
+- A Next.js application with a Postgres database, packaged as a Cloudron app (Docker image + `CloudronManifest.json`).
 - Uses Cloudron's **postgresql + sendmail + localstorage** addons.
-  Cloudron's built-in container supervisor keeps it running.
-- Sign-in changes from 4-digit email **codes** to email **links**
-  ("magic link" authentication). Users click a link in their inbox
-  instead of typing a code.
-- One visible process, one port (3000), one health check
-  (`/api/health`), ~512 MiB memory — the same footprint as the
-  existing app.
+- Cloudron's built-in container supervisor keeps it running.
+- Sign-in changes from 4-digit email **codes** to email **links** ("magic link" authentication). Users click a link in their inbox instead of typing a code.
+- One visible process, one port (3000), one health check (`/api/health`), ~512 MiB memory the same footprint as the existing app.
 
 ## What does NOT carry over
 
 - **No user accounts.** New sign-ins start fresh.
-- **No published rules from the old database.** We'll count the
-  existing `rules` table before cutover and decide whether to publish
-  a read-only archive (CSV/JSON) somewhere for anyone looking for
-  their old work.
+- **No published rules from the old database.** We'll count the existing `rules` table before cutover and decide whether to publish a read-only archive (CSV/JSON) somewhere for anyone looking for their old work.
 - **No chatbot.**
 
 ## How the cutover will work
@@ -43,30 +32,22 @@ Side-by-side, the legacy app keeps running untouched
 until the new one is verified.
 
 1. **Staging phase.** New app installed at
-   `staging.communityrule.info` (auto-provisioned by Cloudron). Legacy
-   app at the apex is not touched. Quiet testing within MEDLab/stakeholders.
-2. **Cutover phase.** When staging is green and we're ready, schedule
-   a low-traffic window. During the window (roughly 5–15 minutes of
-   apex downtime):
+   `staging.communityrule.info` (auto-provisioned by Cloudron). Legacy app at the apex is not touched. Quiet testing within MEDLab/stakeholders.
+2. **Cutover phase.** When staging is green and we're ready, schedule a low-traffic window. During the window (roughly 5–15 minutes of apex downtime):
    - Take a final backup of the legacy app (Cloudron one-click).
-   - Pull a copy of the legacy `rules` table if we decided to publish
-     an archive.
+   - Pull a copy of the legacy `rules` table if we decided to publish an archive.
    - Uninstall the legacy app at the apex `communityrule.info`.
    - Move the new app to the apex.
    - Smoke-test, confirm backups are on, done.
-3. **Post-cutover.** Legacy backup retained ≥ 90 days as a safety net.
-   Legacy source repos get README pointers to the new app and are archived.
+3. **Post-cutover.** Legacy backup retained ≥ 90 days as a safety net. Legacy source repos get README pointers to the new app and are archived.
 
-Rollback plan during the window: restore the legacy backup to a scratch
-Cloudron slot and point DNS back. Realistic only if we discover
-something genuinely broken in the first few minutes.
+Rollback plan during the window: restore the legacy backup to a scratch Cloudron slot and point DNS back. Realistic only if we discover something genuinely broken in the first few minutes.
 
 ## Rough timeline
 
 Roughly this order:
 
-1. **Code prep** — small local change so the app reads Cloudron's
-   injected `CLOUDRON_*` env vars natively. No infra impact.
+1. **Code prep** — small local change so the app reads Cloudron's injected `CLOUDRON_*` env vars natively. No infra impact.
 2. **Build and push the app image** to a container registry.
 3. **Install at staging** subdomain, smoke test, soft launch.
 4. **Apex cutover window** — the brief downtime above.