[prompt-analysis] Copilot PR Prompt Analysis - 2026-07-10

[github-actions[bot]]

📊 Summary

Analysis Period: Last 30 days
Total PRs: 1,000 | Merged: 795 (80.3%) | Closed: 200 (20.1%) | Open: 5

Prompt Categories and Success Rates

Category	Total	Merged	Closed	Success Rate
Refactor	87	70	16	81%
Bug Fix	668	533	131	80%
Testing	82	65	17	79%
Docs	113	90	23	80%
Feature	34	26	8	76%
Other	16	11	5	69%

✅ Successful Prompt Patterns

Common characteristics in merged PRs:

• Avg body length: 183 words
• Top action verbs: improve (92%), update (90%), fix (85%)
• WIP label rate: only 1.1% of merged PRs were marked [WIP]
• File references in body: 27% include specific file paths

Example merged prompts:

View examples

1. PR Fix Daily Elixir Credo Snippet Audit bootstrap step for Mix compatibility #44762 — Fix Daily Elixir Credo Snippet Audit bootstrap step for Mix → Merged

   "The Daily Elixir Credo Snippet Audit workflow was failing in its setup phase before any snippet lint..."

2. PR feat: propagate OAuth token check failure to conclusion job failure issue #44756 — feat: propagate OAuth token check failure to conclusion job → Merged

   "When a workflow is configured with an OAuth token (gho_...), the activation job fails at `check-oa..."

3. PR docs(parser): remove phantom function from README Public API table #44739 — docs(parser): remove phantom function from README Public API → Merged

   "pkg/parser/README.md documented LocateJSONPathInYAMLWithAdditionalProperties(...) which no longer exists..."

❌ Unsuccessful Prompt Patterns

Common characteristics in closed PRs:

• Avg body length: 180 words (similar to merged — length alone doesn't predict success)
• WIP label rate: 15% of closed PRs were marked [WIP] (vs 1.1% merged)
• Incomplete checklists: 12 closed PRs had unchecked - [ ] task items
• Lower-success verbs: bump (68%), chore (67%), docs (71%)

Example closed prompts:

View examples

1. PR chore: bump @playwright/cli default version from 0.1.15 to 0.1.17 #44719 — chore: bump @playwright/cli default version → Closed

   "Bumps DefaultPlaywrightCLIVersion from 0.1.15 to 0.1.17..."

2. PR [WIP] Investigate recurring Smoke Antigravity failures #44716 — [WIP] Investigate recurring Smoke Antigravity failures → Closed

   "- [x] Investigated root cause... - [ ] Apply fix - [ ] Verify tests pass..."

3. PR [WIP] Improve test quality in pkg/actionpins/spec_test.go #44707 — [WIP] Improve test quality in pkg/actionpins/spec_test.go → Closed

   "- [ ] Add TestSpec_PublicAPI_ResolveActionPin_DynamicReturnsEmptySHA..."

🔑 Key Insights

• [WIP] is a strong failure signal: 15% of closed PRs used [WIP] labels vs only 1.1% of merged PRs — these are exploratory or incomplete drafts that rarely land.
• Incomplete checklists predict closure: PRs with unchecked - [ ] task items were disproportionately closed; all-checked checklists correlate with merges.
• Prompt length doesn't matter: Merged (183 words) and closed (180 words) PR bodies are nearly identical in length — quality and specificity matter more than verbosity.
• Refactoring PRs lead success rates at 81%: Clear, scoped refactors with specific targets are the most reliably merged category.

💡 Recommendations

1. DO: Use precise action verbs like improve, update, or fix with a specific target (file, function, behavior)
2. DO: Complete all checklist items before submitting — or omit checklists if the work is already done
3. DO: Favor refactor and fix framings over chore/bump for higher merge rates
4. AVOID: Submitting PRs with [WIP] in the title — they have a 15x higher closure rate
5. AVOID: Vague scope words like "improve generally" or "investigate" — these correlate with closed PRs

📈 Historical Trends (7-day)

Date	PRs	Merged	Success Rate	Top Category
2026-07-10	1000	795	80.3%	Bug Fix
2026-07-08	1000	795	80.3%	Bug Fix
2026-07-07	1000	792	79.8%	Bug Fix
2026-07-06	1000	793	80.4%	Bug Fix / Feature
2026-07-05	1000	805	80.8%	Bug Fix
2026-07-04	1000	798	81.1%	Bug Fix
2026-07-03	1000	801	81.8%	Bug Fix

Trend: Overall success rate is stable in the 79–82% range. Bug Fix continues to dominate by volume. Today's slight dip (80.3%) is within normal variance — no concerning patterns.

---

References:

• §29123156221

Generated by 🔍 Copilot PR Prompt Pattern Analysis · 27.8 AIC · ⌖ 12.4 AIC · ⊞ 7.4K · ◷

• expires on Jul 11, 2026, 1:04 PM UTC-08:00

[github-actions[bot]]

This discussion has been marked as outdated by Copilot PR Prompt Pattern Analysis.

A newer discussion is available at Discussion #45010.