Summary
A daily review of agentic workflow prompts active in the last 24 hours found several report-generating workflows that don't follow the project's markdown style guidelines (header hierarchy, progressive disclosure, and clear report structure). Consolidating findings here per the Workflow Normalizer process.
Workflows Reviewed: 91 (workflows active in the last 24h with a matching .md file)
Report-Generating Workflows Analyzed: 56 (those using create-issue, create-discussion, or add-comment safe-outputs)
Non-Compliant: 6 flagged below (highest-impact / clearest gaps; several smoke-test workflows were intentionally excluded as low-value for style normalization)
Non-Compliant Workflows
| Workflow File |
Issue Found |
.github/workflows/daily-harness-experiment-proposer.md |
Uses #/## top-level headers instead of h3+ (###); no <details>/progressive disclosure guidance for long content |
.github/workflows/code-scanning-fixer.md |
Uses #/## headers; no progressive disclosure guidance; no report structure notes |
.github/workflows/firewall-escape.md |
Uses #/## headers for report/test sections; no <details> collapsing guidance despite likely long test output |
.github/workflows/issue-arborist.md |
Uses #/## headers; no progressive disclosure guidance for created issue content |
.github/workflows/copilot-centralization-optimizer.md |
Uses #/## headers; no progressive disclosure or report-structure guidance |
.github/workflows/step-name-alignment.md |
Uses #/## headers throughout; no progressive disclosure guidance |
Required Changes (per workflow)
For each workflow above:
- Header levels: Update prompt instructions/examples so any generated report content starts at
### (h3) or lower — reserve #/## for the workflow's own internal prompt section titles only, and explicitly instruct the agent to use ###+ in the issue/comment/discussion body it produces.
- Progressive disclosure: Add explicit instructions to wrap long or detailed content (logs, lists, diffs, extended findings) in
<details><summary><b>...</b></summary> ... </details> blocks so the top-level report stays scannable.
- Report structure: Add a short "Output Format" section (or equivalent) showing the expected structure — key metrics/status up top, details collapsed below, and a clear "Next Steps" or checklist section using
- [ ] syntax where applicable.
Example: Progressive Disclosure Pattern
### Findings Summary
**Issues found**: 3 | **Severity**: Medium
<details>
<summary><b>View detailed findings</b></summary>
1. Finding one details...
2. Finding two details...
3. Finding three details...
</details>
### Next Steps
- [ ] Review findings
- [ ] Apply fixes
Good Reference Examples
.github/workflows/daily-repo-chronicle.md (if present) and audit-workflows-style prompts follow this pattern well: concise top-level summary, <details> for full breakdowns, checklists for next steps.
- Compliant examples found in this review:
pr-sous-chef.md, issue-monster.md, auto-triage-issues.md, duplicate-code-detector.md, semantic-function-refactor.md.
Design Principles Applied
- Trust through clarity: key results visible immediately, details opt-in via
<details>.
- User-focused: readers can decide in seconds whether to dig deeper.
- Consistent experience: same formatting pattern across all workflow reports.
Next Steps
Generated by 🔧 Workflow Normalizer · copilot · auto · 27.9 AIC · ⌖ 5.97 AIC · ⊞ 9.5K · ◷
Summary
A daily review of agentic workflow prompts active in the last 24 hours found several report-generating workflows that don't follow the project's markdown style guidelines (header hierarchy, progressive disclosure, and clear report structure). Consolidating findings here per the Workflow Normalizer process.
Workflows Reviewed: 91 (workflows active in the last 24h with a matching
.mdfile)Report-Generating Workflows Analyzed: 56 (those using
create-issue,create-discussion, oradd-commentsafe-outputs)Non-Compliant: 6 flagged below (highest-impact / clearest gaps; several smoke-test workflows were intentionally excluded as low-value for style normalization)
Non-Compliant Workflows
.github/workflows/daily-harness-experiment-proposer.md#/##top-level headers instead of h3+ (###); no<details>/progressive disclosure guidance for long content.github/workflows/code-scanning-fixer.md#/##headers; no progressive disclosure guidance; no report structure notes.github/workflows/firewall-escape.md#/##headers for report/test sections; no<details>collapsing guidance despite likely long test output.github/workflows/issue-arborist.md#/##headers; no progressive disclosure guidance for created issue content.github/workflows/copilot-centralization-optimizer.md#/##headers; no progressive disclosure or report-structure guidance.github/workflows/step-name-alignment.md#/##headers throughout; no progressive disclosure guidanceRequired Changes (per workflow)
For each workflow above:
###(h3) or lower — reserve#/##for the workflow's own internal prompt section titles only, and explicitly instruct the agent to use###+ in the issue/comment/discussion body it produces.<details><summary><b>...</b></summary> ... </details>blocks so the top-level report stays scannable.- [ ]syntax where applicable.Example: Progressive Disclosure Pattern
Good Reference Examples
.github/workflows/daily-repo-chronicle.md(if present) andaudit-workflows-style prompts follow this pattern well: concise top-level summary,<details>for full breakdowns, checklists for next steps.pr-sous-chef.md,issue-monster.md,auto-triage-issues.md,duplicate-code-detector.md,semantic-function-refactor.md.Design Principles Applied
<details>.Next Steps
daily-harness-experiment-proposer.mdwith header/<details>guidancecode-scanning-fixer.mdwith header/<details>guidancefirewall-escape.mdwith header/<details>guidanceissue-arborist.mdwith header/<details>guidancecopilot-centralization-optimizer.mdwith header/<details>guidancestep-name-alignment.mdwith header/<details>guidancemake recompile) after edits