Skip to content

[curriculum-eval] side-quest-11-06-anthropic-key.md: cognitive_load — redundant secret-name-variant list inflates concept count #3406

Description

@github-actions

File: workshop/side-quest-11-06-anthropic-key.md
Overall Score: 5.39 / 10.0 (corpus mean: 6.14)

Flagged Dimensions:

Dimension Score Benchmark Delta
cognitive_load ~3.9 (concept component low; 26 new concepts vs ≤15 ideal) ≤ 15 new concepts large
active_learning 3.5 density ≥ 3 → 10 -6.5

Root Cause (≤ 2 sentences):
Despite a short 665-word body, the page introduces 26 distinct bolded/code terms (ANTHROPIC_API_KEY, engine: claude, api.anthropic.com, ANTHROPIC-API-KEY, CLAUDE_API_KEY, network.allowed, copilot-requests: write, plus every table cell and code-fence token), driven largely by an exhaustive "common mistakes" list of secret-name variants that adds vocabulary without adding a new procedural step. missing_checkpoint is again a false positive from the shortcode regex bug — the page has a real 5-item checkpoint at line 109.

Evidence (quoted from the file):

"Wrong name: any variation (anthropic_api_key, ANTHROPIC-API-KEY, CLAUDE_API_KEY) will cause a silent auth failure... Copied with extra whitespace... Closed the Anthropic tab before saving... Network allow-list missing..."

Learning Science Rationale:
Sweller's Cognitive Load Theory predicts that enumerating multiple near-identical failure variants (four bolded "common mistakes," each introducing its own bolded term) increases extraneous load without adding to the learner's actual procedural skill, since the underlying fix ("use the exact secret name") is a single idea repeated in four surface forms — this is the kind of redundancy the theory specifically flags as harmful to schema construction.

Improvement Prompt (for an agent):

Edit workshop/side-quest-11-06-anthropic-key.md to reduce concept density. Specifically:
1. Collapse the "Common mistakes with this secret" details block from four separately-bolded bullet items into a single short bulleted list under one heading term (e.g. "Common mistakes") without bolding each variant name as its own concept — keep the substance but stop treating each incorrect spelling as a distinct new term.
2. In the "Update your workflow frontmatter" code block, keep only the fields the learner will actually change (engine, network.allowed) and move the note about removing copilot-requests: write into a single inline sentence rather than a separate callout-style aside.
3. Consider adding one short hands-on check between "Get an Anthropic API key" and "Store the key as a repository secret" (e.g. a one-line command to confirm the key format starts with sk-ant-) to raise activity_density.
Re-run the corpus scoring script and confirm new_concepts drops toward 15-18 while checklist_items and activity_density are unchanged or improved.

Expected Score After Fix: 6.2 / 10.0

Generated by 🔬 Curriculum Quality Evaluator · copilot · auto · 76.7 AIC · ⌖ 17.4 AIC · ⊞ 9.5K ·

  • expires on Sep 16, 2026, 1:00 PM UTC

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions