[curriculum-eval] side-quest-11-06-anthropic-key.md: cognitive_load — wrong-name variants inflate concept count

[github-actions]

File: workshop/side-quest-11-06-anthropic-key.md
Overall Score: 5.39 / 10.0 (corpus mean: 6.15)

Flagged Dimensions:

Dimension	Score	Benchmark	Delta
cognitive_load	7.2	≤15 new concepts	26 new concepts (+11 over)
active_learning	3.5	density ≥3	1.05
checkpoint_quality	0.0*	≥4 items	*tooling false-negative (see corpus scorecard)

Root Cause (≤ 2 sentences):
At 665 words — comfortably under the length guideline — this is an operational how-to (get a key, store a secret, edit frontmatter, compile) that nonetheless surfaces 26 distinct inline-code/bold terms because every UI label (Settings, Secrets and variables, Actions, New repository secret), every frontmatter key (engine: claude, network.allowed, ANTHROPIC_API_KEY), and every common-mistake variant name (anthropic_api_key, ANTHROPIC-API-KEY, CLAUDE_API_KEY) is rendered as its own inline-code span. The page has almost no active-learning component beyond following numbered instructions — it reads as a checklist walkthrough, not an exercise, so activity density stays at 1.05 despite 5 checklist items.

Evidence (quoted from the file):

Wrong name: any variation (anthropic_api_key, ANTHROPIC-API-KEY, CLAUDE_API_KEY) will cause a silent auth failure. The name must be exactly ANTHROPIC_API_KEY.

Learning Science Rationale:
Sweller's Cognitive Load Theory flags this pattern of listing every wrong-name variant as its own code span: while pedagogically useful as an error-prevention aid, each variant is counted (and processed) as a distinct new term, needlessly inflating extraneous load for what is fundamentally one instruction ("use this exact string"). Consolidating the "common mistakes" list into fewer, non-code-formatted prose bullets — since these are illustrative wrong answers, not vocabulary to learn — would reduce apparent concept load without losing the warning's value.

Improvement Prompt (for an agent):

Open workshop/side-quest-11-06-anthropic-key.md. This page surfaces 26 distinct
inline-code/bold terms in only 665 words (ideal ≤15), largely because the
"Common mistakes with this secret" collapsible renders every wrong-name
variant (anthropic_api_key, ANTHROPIC-API-KEY, CLAUDE_API_KEY) as its own
inline-code span alongside every UI label and frontmatter key already used
elsewhere on the page.

1. In the "Common mistakes with this secret" `<details>` block, rewrite the
   "Wrong name" bullet to describe the failure mode in prose without wrapping
   each incorrect variant in backticks (e.g. "any lowercase, hyphenated, or
   differently-named variation will cause a silent auth failure — the name
   must be exactly `ANTHROPIC_API_KEY`"), keeping only the one correct code
   span.
2. Add one short active checkpoint before the final Checkpoint section — for
   example a `<details>` "Quick check" asking the learner to predict what
   happens if they name the secret `Anthropic_Api_Key` instead of
   `ANTHROPIC_API_KEY`, with the answer revealed — to raise activity density
   above 3.0 and turn passive reading into a predict-then-verify exercise.
3. Do not remove any of the existing setup steps, the frontmatter example, or
   the compile step — they are load-bearing procedural content.
4. Run `npx --yes markdownlint-cli2 "**/*.md"` after editing and fix any lint
   errors before finishing.

Expected Score After Fix: 6.3 / 10.0

Generated by 🔬 Curriculum Quality Evaluator · copilot · auto · 142.5 AIC · ⌖ 18.6 AIC · ⊞ 9.5K · ◷

• expires on Sep 11, 2026, 3:49 PM UTC