[prompt-analysis] Copilot PR Prompt Analysis - 2026-10-16

[github-actions[bot]]

🤖 Copilot PR Prompt Pattern Analysis - 2026-10-16

Summary

Analysis Period: Last 30 days
Total PRs: 606 | Merged: 501 (82.7%) | Closed: 94 (15.5%) | Open: 11 (1.8%). Merge rate of completed PRs: 84.2%. No empty PR bodies.

Prompt Categories and Success Rates (by title keyword, completed PRs only)

Category	Total	Merged	Success Rate
Bug Fix	73	65	90%
Feature Addition	102	83	83%
Refactoring	11	9	82%
Documentation	27	22	81%
Testing	8	6	75%
Other/uncategorized	385	316	84%

Prompt Analysis

✅ Successful Patterns

• Average body length: ~165 words (median 107).
• Action verbs: Fix, Add, Support, Respect, Correct.
• PRs whose body references specific files (.go/.cjs/.md/.yml): 94 merged vs 16 closed (85% merge rate); without file references: 407 vs 78 (84%) — little difference.

Examples:

1. Support create-issue: false to disable automatic issue output #69369: Support create-issue: false to disable automatic issue output → Merged
2. Fix consumer-broken link in agentic-workflows skill #69365: Fix consumer-broken link in agentic-workflows skill → Merged
3. Correct table-ledger payloads behind recent workflow audit failures #69355: Correct table-ledger payloads behind recent workflow audit failures → Merged

❌ Unsuccessful Patterns

• Average body length: ~202 words, longer than merged PRs (165).
• Closed PRs skew toward broader, multi-part tasks (audits, sync, cross-cutting changes).

Examples:

1. Sync gh-aw-actions pre-release status with gh-aw #69367: Sync gh-aw-actions pre-release status with gh-aw → Closed
2. Fix missing redaction step ID in AWF artifact upload guards #69183: Fix missing redaction step ID in AWF artifact upload guards → Closed
3. Add outcome clustering to workflow audits #68924: Add outcome clustering to workflow audits → Closed

Key Insights

• Bug fixes merge most often (90%); testing-only prompts are weakest (75%, small sample of 8).
• Shorter prompts correlate with merges (165 vs 202 words on average).
• File references showed no meaningful success difference this period.

Recommendations

1. DO: Keep prompts focused on a single, well-scoped change.
2. DO: Frame bug fixes with concrete failure symptoms.
3. AVOID: Bundling large cross-cutting or exploratory work into one prompt; split it.

Historical Trends

Date	PRs	Success Rate	Top Category
2026-10-16	606	84.2%	Bug Fix (90%)
2026-10-14	601	85.1%	Bug Fix (89%)

Trend: Stable; merge rate dipped slightly (-0.9pt).

---

Generated by Copilot PR Prompt Analysis (Run: 38964253454)

Generated by 🔍 Copilot PR Prompt Pattern Analysis · copilot · auto · 13.6 AIC · ⌖ 0.713 AIC · ⊞ 11.5K · ◷

• expires on Oct 17, 2026, 12:47 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 #69783.