BOONDOGGLE SCORE

LABOR: Claude Code
CONTEXT REQUIRED: glimmer-consequentialism/generate.js, glimmer-consequentialism/index.html, glimmer-consequentialism/example-game.json, Zebonastic project root

PROMPT:

Deploy the Cobra Effect game to the Zebonastic site.

1. Install the canvas package if not already present:
   npm list canvas || npm install canvas

2. Copy the three game files into the Zebonastic project:
   - glimmer-consequentialism/index.html → keep as source engine
   - glimmer-consequentialism/generate.js → the generator
   - glimmer-consequentialism/example-game.json → use as game.json

3. Run the generator with --validate first:
   node glimmer-consequentialism/generate.js \
     glimmer-consequentialism \
     glimmer-consequentialism/index.html \
     --validate

4. If validation passes, deploy:
   node glimmer-consequentialism/generate.js \
     glimmer-consequentialism \
     glimmer-consequentialism/index.html \
     --nextjs .

5. Verify output:
   ls -la public/games/cobra-effect/
   ls -la public/games/cobra-effect/assets/

   Expected:
   - index.html (with title "The Cobra Effect", description, keywords meta tags)
   - user-config.js (GAME_CONFIG from game.json, no _mechanic_argument)
   - assets/ containing a.png, b.png, c.png, d.png (placeholder PNGs)

6. Confirm user-config.js does NOT contain _mechanic_argument:
   grep "_mechanic_argument" public/games/cobra-effect/user-config.js

   Expected: only the comment line, not an object key.

7. Commit and push:
   git add public/games/cobra-effect/
   git commit -m "feat: add cobra-effect starter template (consequentialism, Goodhart's Law)"
   git push origin main

Report: validation output, directory listing, grep result, commit hash.

EXPECTED OUTPUT: public/games/cobra-effect/ with index.html, user-config.js, and 4 placeholder PNGs. Committed to main.

HANDOFF CONDITION: Directory exists. All four PNGs generated. _mechanic_argument appears only in the comment, not as a GAME_CONFIG key. Commit pushed.

DEPENDENCY: None — deploy infrastructure is proven.

LABOR: Human
SUPERVISORY CAPACITY: [PA] — Plausibility Auditing: after Vercel deploys (1–2 minutes), verify the card and game load correctly.

ACTION:

Check 1 — Card on /games:

• Card labeled "The Cobra Effect" appears
• Description reads correctly (no & — entity fix should be live)
• Tags include "consequentialism" and "ethical play"

Check 2 — Game loads at /games/cobra-effect:

• Intro screen shows: role, context, board note ("Board metric: Satisfaction Score — target 55"), four node descriptions in order A→B→C→D

Check 3 — One full play-through:

• Allocate all 5 tokens to D (Satisfaction Score) every turn for all 12 quarters
• Confirm: next-turn button is disabled until all 5 tokens are placed
• Confirm: Satisfaction Score rises quickly, Staff Morale declines visibly
• Confirm: around turn 9–10, Staff Morale bar turns orange/red, Patient Throughput drops sharply
• Confirm: reveal appears with four sections; "What this game argues" reads: "When a measure becomes a target, it ceases to be a good measure of the thing it was designed to track."
• Confirm: closing question reads: "What was the Satisfaction Score actually measuring?"

FAIL TRIAGE:

• Card missing → index.html missing <title> tag. Check the generated file.
• Game blank → check browser console. Likely user-config.js 404. Check assets path.
• Next-turn button never enables → tokensLeft !== 0 check. Try placing all 5 tokens before checking.
• Staff Morale doesn't collapse → hidden feedback thresholds not triggering. Check that GAME_CONFIG in the loaded page does NOT contain hidden_feedback override (should be engine-default).
• Reveal shows wrong state → open console after reveal, type state to see final values and which state key was selected.

HANDOFF CONDITION: Card visible with correct tags. Game loads. Pure-D strategy produces high_metric_depleted reveal state. Mechanic argument is engine text.

DEPENDENCY: Step 1.

LABOR: Human
SUPERVISORY CAPACITY: [IJ] — Interpretive Judgment: does the mechanic produce the glimmer, or does it feel like a resource management puzzle with a surprise ending?

ACTION:

Play the game three ways and assess:

Run 1 — Pure D optimization (all 5 tokens in Satisfaction Score every turn):
Does the collapse feel discovered or telegraphed? The bar color change on Staff Morale is intentional — a visual signal without an explanation. Does the player feel implicated when Patient Throughput crashes, or does it feel like the game "punished" them?

Run 2 — Balanced (distribute tokens across all four nodes):
Does this produce high_metric_intact? Does the reveal text for that state feel honest — "you found a balance most administrators do not find" — or does it feel like a prize?

Run 3 — Protect A heavily (3–4 tokens in Staff Morale every turn):
Does this produce low_metric_intact? Does the reveal — "you protected the system at the cost of the metric. Whether that was the right trade is not for this screen to say." — actually feel like it's not saying?

The test: after reading the reveal, does the player feel like they received a mirror or a verdict? If it feels like a verdict, the state text needs revision.

HANDOFF CONDITION: At least two of the three reveal states produce the glimmer. If any state text reads as a verdict, note the specific text that needs revision before student use.

DEPENDENCY: Step 2.

LABOR: Human
SUPERVISORY CAPACITY: [EI] — Executive Integration: confirm coherence across the full games system.

ACTION:

1. /games — Evolution of Trust, Veil of Ignorance, and The Cobra Effect all appear. Tag filter "ethical play" returns all three. Tag filter "consequentialism" returns only Cobra Effect.
2. /games/cobra-effect — game plays correctly on fresh browser (cleared cache).
3. Both placeholder PNGs (node A–D) are visible in the assignment phase as neutral gray squares with node key labels.
4. Course taxonomy check: Week 2 of the Ethical Play syllabus covers "Consequentialism as Mechanic — the VE causal chain display." The keywords in the deployed game (consequentialism, goodharts law, metrics, measurement, ethical play) match what a student in Week 2 would search for.

HANDOFF CONDITION: All three games visible and tagged correctly. Cobra Effect keywords match Week 2 course vocabulary. Released.

DEPENDENCY: Step 3.