[prompt-analysis] Copilot PR Prompt Analysis - 2026-10-19

[github-actions[bot]]

🤖 Copilot PR Prompt Pattern Analysis - 2026-10-19

Summary

Analysis Period: Last 30 days
Total PRs: 611 | Merged: 506 (82.8%) | Closed: 99 (16.2%) | Open: 6

Merge rate among completed PRs: 83.6%. No PRs had empty bodies.

Prompt Categories and Success Rates

(Categorized by title keywords; success = merged / (merged + closed))

Category	Total	Merged	Success Rate
Bug Fix	77	68	88.3%
Feature Addition	90	71	78.9%
Refactoring	17	13	76.5%
Documentation	31	25	80.6%
Testing	21	18	85.7%
Other	375	311	83.8%

Prompt Analysis

✅ Successful Prompt Patterns

• Average prompt (body) length: 162 words
• Bug-fix and test-oriented prompts perform best.
• Prompts referencing specific files/extensions: 90 merged / 17 closed (84%).

Examples:

1. PR Preserve AWF model inventories with null endpoints and live reasoning metadata (G-7) #70013: Preserve AWF model inventories with null endpoints and live reasoning metadata → Merged
2. PR Recognize merged base ancestors when the base tip advances #69998: Recognize merged base ancestors when the base tip advances → Merged
3. PR Prevent OTLP placeholder masks from suppressing boolean job outputs #69996: Prevent OTLP placeholder masks from suppressing boolean job outputs → Merged

❌ Unsuccessful Prompt Patterns

• Average prompt length: 200 words (about 24% longer than merged).
• Closed PRs skew toward features, refactors, and broad accounting/cleanup changes.

Examples:

1. PR Use artifact-backed credit accounting for daily code metrics #69988: Use artifact-backed credit accounting for daily code metrics → Closed
2. PR Fix cached audit ledger schema fixture causing CI failure #69940: Fix cached audit ledger schema fixture causing CI failure → Closed
3. PR Add opt-in duplicate PR cleanup to update and upgrade #69721: Add opt-in duplicate PR cleanup to update and upgrade → Closed

Key Insights

• Specificity: Bug fix prompts (88%) beat feature (79%) and refactor (77%) prompts.
• Length: Closed PR prompts are longer on average (200 vs 162 words), suggesting sprawling scope hurts.
• Caveat: Small categories (refactor n=17, test n=21) are weak statistically.

Recommendations

1. DO: Frame tasks as a concrete problem with clear expected behavior.
2. DO: Keep prompts focused and name relevant files.
3. AVOID: Bundling features, refactors, and cleanup in one prompt; split them.

Historical Trends

Date	PRs	Success Rate	Top Category
2026-10-19	611	83.6%	Bug Fix (88%)
2026-10-18	602	83.6%	Bug Fix (86%)

Trend: Stable merge rate, slight volume growth.

---

Generated by Copilot PR Prompt Analysis (Run: 39321668556)

Generated by 🔍 Copilot PR Prompt Pattern Analysis · copilot · auto · 16 AIC · ⌖ 0.714 AIC · ⊞ 11.5K · ◷

• expires on Oct 20, 2026, 12:48 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 #70629.