Normalize report-format guidance across non-compliant agentic workflows

.github/workflows/auto-triage-issues.md

@@ -207,7 +207,7 @@ For the triggering issue (on issue events), you can omit `item_number`:
 }
 ```
 
-## Scheduled Run Report
+### Scheduled Run Report
 
 When running on schedule, create a discussion report following these formatting guidelines:
 

.github/workflows/daily-syntax-error-quality.md

@@ -114,7 +114,7 @@ cat /tmp/gh-aw/agent/candidates.txt
 1. Simple workflow (< 100 lines, minimal config)
 2. Complex workflow (> 300 lines, many tools/features)
 
-## Phase 2: Generate Syntax Errors
+### Phase 2: Generate Syntax Errors
 
 For each selected workflow, create exactly **1 test case** with a different error type:
 
@@ -177,7 +177,7 @@ For each workflow:
    }
    ```
 
-## Phase 3: Run Compiler and Capture Output
+### Phase 3: Run Compiler and Capture Output
 
 For each test case:
 
@@ -225,7 +225,7 @@ test-1 | <workflow> | <error type> | clarity:<n>/25 actionability:<n>/25 context
 
 Collect key strengths (1–2 bullets) and improvement suggestions (1–2 bullets) per test. Do **not** reproduce the full compiler output in your report — reference file:line only.
 
-## Phase 6: Create Issue with Suggestions
+### Phase 6: Create Issue with Suggestions
 
 **Only create an issue if**:
 - Average score < 65 across all test cases, OR

.github/workflows/designer-drift-audit.md

@@ -176,13 +176,16 @@ If **drift is detected**, create an issue with:
 **Body structure**:
 
 ```
-## Designer Drift Report — <date>
+### Designer Drift Report — <date>
 
 ### Summary
 <1-2 sentence overview: X drift items found across Y categories>
 
 ### Drift Items
 
+<details>
+<summary><b>View Drift Items by Category</b></summary>
+
 #### <Category> (e.g., "Missing Safe Output Types")
 - **Source**: `<reference-doc-path>`
 - **Missing from**: `SKILL.md` / `agent file` / both
@@ -191,6 +194,8 @@ If **drift is detected**, create an issue with:
 
 ...repeat for each drift item...
 
+</details>
+
 ### No Drift
 <list categories that passed with ✅>
 

.github/workflows/draft-pr-cleanup.md

@@ -208,7 +208,7 @@ Create a summary of actions taken:
 *Draft PR Cleanup workflow run: ${{ github.run_id }}*
 ```
 
-## Important Guidelines
+### Important Guidelines
 
 ### Fair and Transparent
 - Calculate inactivity objectively based on measurable activity
@@ -234,7 +234,7 @@ Create a summary of actions taken:
 - **PR with `stale-draft` but recent activity**: Remove `stale-draft` label if activity < 10 days
 - **Bot-created PRs**: Apply same rules, but consider if bot is still active
 
-## Success Metrics
+### Success Metrics
 
 Effectiveness measured by:
 - **Draft PR rate**: Reduce from 9.6% to <5% over time

.github/workflows/mattpocock-skills-reviewer.md

@@ -251,15 +251,15 @@ Applied **`/tdd`** and **`/zoom-out`** — requesting changes on test coverage g
 If the review is complex or the overall findings are significant, post a single `add-comment` with a concise summary for the author. Apply progressive disclosure: one-line outcome visible, details in `<details>` blocks.
 Use `###` or lower for any headers — never `#` or `##`.
 
-## Scope Rules
+### Scope Rules
 
 - **Review changed lines only** — do not critique unchanged code
 - **Prioritise impact** — security > correctness > maintainability > style
 - **Maximum 10 inline comments** — pick the highest-value issues
 - **Skip auto-generated files** — lock files, generated code, build artifacts
 - **Be constructive** — suggest improvements, not just problems
 
-## Tone
+### Tone
 
 - Professional and collegial — not grumpy, not sycophantic
 - Reference skills by name so the author can learn more