[prompt-analysis] Copilot PR Prompt Analysis - 2026-10-08

[github-actions[bot]]

🤖 Copilot PR Prompt Pattern Analysis - 2026-10-08

Summary

Analysis Period: Last 30 days
Total PRs: 552 | Merged: 463 (83.9%) | Closed: 86 (15.6%) | Open: 3

No PRs had empty bodies. Categories are inferred from title keywords (the "Other" bucket is large, so treat category rates as indicative).

Prompt Categories and Success Rates

Category	Total (completed)	Merged	Success Rate
Bug Fix	90	78	86.7%
Feature Addition	82	64	78.0%
Refactoring	14	12	85.7%
Documentation	19	15	78.9%
Testing	9	7	77.8%
Other	335	287	85.7%

Prompt Analysis

✅ Successful Prompt Patterns

• Average prompt length: 163.6 words (median 109)
• Specific file references (.go/.cjs/.md/.ts/.yml) appear in 91 merged PRs vs 16 closed PRs
• Action verbs: fix, add, support, normalize, prevent

Examples:

1. PR Normalize blank optional safe-output fields before validation #66724: Normalize blank optional safe-output fields before validation → Merged
2. PR Support model routing across Claude, Codex, and pi #66723: Support model routing across Claude, Codex, and pi → Merged

❌ Unsuccessful Prompt Patterns

• Average prompt length: 184.7 words (median 98) – closed PRs have a longer tail of verbose prompts
• Closed PRs skew toward narrow-environment or edge-case issues and overlapping/duplicate fixes

Examples:

1. PR Avoid GHES masking of cross-job workflow outputs #66678: Avoid GHES masking of cross-job workflow outputs → Closed
2. PR Prevent unresolved AI credit estimates from blocking workflows #66559: Prevent unresolved AI credit estimates from blocking workflows → Closed
3. PR Handle models missing from the pricing table #66552: Handle models missing from the pricing table → Closed

Key Insights

• Bug fixes have the highest merge rate (86.7%); feature prompts lag (78%), likely due to scope ambiguity.
• Closed PRs have a higher mean but lower median length: a few very long, sprawling prompts hurt.
• The three closed examples above overlap in topic (pricing/credit estimation), suggesting duplicate attempts at the same problem.

Recommendations

1. DO: Reference specific files/functions and expected behavior.
2. DO: Keep feature prompts tightly scoped with acceptance criteria.
3. AVOID: Opening parallel PRs for the same issue, and very long unfocused prompts.

Historical Trends

Date	PRs	Merge Rate	Top Category
2026-10-08	552	83.9%	Bug Fix (86.7%)
2026-10-07	543	83.3%	Testing (89%)

Trend: Merge rate is stable (~83–84%); bug fix rate steady at ~86%.

---

Generated by Copilot PR Prompt Analysis (Run: 37842005984)

Generated by 🔍 Copilot PR Prompt Pattern Analysis · copilot · auto · 13.3 AIC · ⌖ 0.687 AIC · ⊞ 10.9K · ◷

• expires on Oct 9, 2026, 12:51 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 #67276.