[prompt-analysis] Copilot PR Prompt Analysis - 2026-05-14

[github-actions[bot]]

Summary

Analysis Period: Last 30 days (2026-04-26 to 2026-05-14)
Total PRs: 1,000 | Merged: 792 (79.4%) | Closed: 206 (20.6%) | Open: 2

Overall, Copilot PRs have a strong 79.4% merge success rate across 1,000 PRs analyzed.

Prompt Categories and Success Rates

Category	Total	Merged	Closed	Success Rate
Bug Fix	800	637	162	80%
Feature Addition	127	95	31	75%
Refactoring	29	24	5	83%
Documentation	19	17	2	89%
Testing	9	8	1	89%
Other	16	11	5	69%

Prompt Analysis

✅ Successful Prompt Patterns

Conventional commit prefixes with highest success rates:

Prefix	Total	Merged	Success Rate
remove	12	12	100%
test	4	4	100%
improve	9	8	89%
refactor	40	34	85%
add	90	74	82%
fix	251	205	82%
feat	102	83	81%

Example successful PR titles:

1. PR fix: include effective tokens count in PR comment generated footer #32215: fix: include effective tokens count in PR comment generated footer → Merged
2. PR Add emoji frontmatter field #32200: Add emoji frontmatter field → Merged
3. PR Add observability.otlp.ignore-if-missing to downgrade missing OTLP config to warnings #32173: Add observability.otlp.ignore-if-missing to downgrade missing OTLP config to warnings → Merged

❌ Unsuccessful Prompt Patterns

Common patterns in closed PRs:

Pattern	Total	Merged	Success Rate
[WIP] prefix	29	7	24%
chore prefix	16	9	56%
No conventional prefix	398	304	76%

Example closed PR titles:

1. PR [WIP] Fix failing GitHub Actions job 'test' #32190: [WIP] Fix failing GitHub Actions job 'test' → Closed
2. PR Enable shared OTLP reuse across all agentic workflows #32158: Enable shared OTLP reuse across all agentic workflows → Closed (no prefix)
3. PR Raise Daily Cache Strategy Analyzer ET budget above default 25M cap #32209: Raise Daily Cache Strategy Analyzer ET budget above default 25M cap → Closed

Key Insights

• [WIP] prefix strongly predicts failure: PRs tagged [WIP] have only a 24% merge rate (7/29). These are likely exploratory attempts that get superseded or abandoned. Avoid creating PRs with [WIP] in the title — complete the implementation before opening the PR.
• Documentation and test prompts are most reliable: Both docs and test categories achieve 89% success rates, suggesting these scoped, well-defined tasks are well-suited for Copilot automation.
• Bug fixes dominate volume but trail in success: Bug fix prompts account for 80% of all PRs (800/1000) but have a slightly lower success rate (80%) than refactoring (83%) or docs (89%), suggesting more complex scoping challenges for bug-fix tasks.

Recommendations

Based on today's analysis:

1. DO: Use specific conventional commit prefixes (fix:, feat:, refactor:, add, remove) — they correlate with 80–100% success rates
2. DO: Prefer documentation and test prompts for reliable outcomes — they have the highest success rates (89%) with minimal overhead
3. AVOID: Using [WIP] prefix — it signals incomplete work and has only 24% merge rate; instead, wait until the implementation is ready before opening a PR
4. AVOID: chore prefix — only 56% success rate, possibly because chore tasks are vague or less impactful
5. DO: Include specific, scoped descriptions — successful PRs like fix: include effective tokens count in PR comment generated footer are precise and targeted

Historical Trends

First analysis run — no prior data to compare. Future runs will track trends.

---

References:

• §25885416796

Generated by 🔍 Copilot PR Prompt Pattern Analysis · ● 4.4M · ◷

• expires on May 15, 2026, 9:05 PM UTC

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-05-15T21:05:22.422Z.

Closed by Workflow