[prompt-analysis] Copilot PR Prompt Analysis - 2026-10-26

[github-actions[bot]]

🤖 Copilot PR Prompt Pattern Analysis - 2026-10-26

Summary

Analysis Period: Last 30 days
Total PRs: 602 | Merged: 503 (83.6%) | Closed: 93 (15.4%) | Open: 6

All PRs had non-empty bodies. Categories are by title keyword matching (heuristic).

Prompt Categories and Success Rates

Category	Total (completed)	Merged	Success Rate
Bug Fix	69	61	88.4%
Feature Addition	81	67	82.7%
Refactoring	10	6	60.0%
Documentation	28	25	89.3%
Testing	14	11	78.6%
Other/uncategorized	394	333	84.5%

Prompt Analysis

✅ Successful Patterns

• Average body length (merged): 161 words
• Narrow, concrete scopes: bug fixes and docs merge at ~88-89%
• Examples: Upgrade Copilot CLI to 1.0.103 with reviewed checksum pin #72099 Upgrade Copilot CLI to 1.0.103 with reviewed checksum pin; Trust same-repository bot PRs triggered by labels #72074 Trust same-repository bot PRs triggered by labels; Bump MCPG to v0.5.0 and firewall to v0.29.0 #71957 Bump MCPG to v0.5.0 and firewall to v0.29.0 → all Merged

❌ Unsuccessful Patterns

• Average body length (closed): 139 words
• Broad or cross-cutting changes (refactors 60%, tests 79%) and security/egress policy changes closed more often
• Examples: Separate Agy bootstrap downloads from runtime firewall egress #72102 Separate Agy bootstrap downloads from runtime firewall egress; Refactor long workflow and CLI functions into focused helpers #71890 Refactor long workflow and CLI functions into focused helpers; Handle ignored errors and discarded results #71892 Handle ignored errors and discarded results → all Closed

Key Insights

• Refactoring is the weakest category (60%, n=10) versus 84% overall; small sample, so treat as directional.
• Bug fix and docs prompts are the most reliable (88-89%).
• Closed PR bodies are slightly shorter than merged ones (139 vs 161 words), suggesting more context helps modestly.

Recommendations

1. DO: Scope prompts to a single well-defined change with explicit version/file targets.
2. DO: Include context and acceptance criteria in the prompt.
3. AVOID: Broad "refactor many functions" or multi-area security-policy prompts; split into smaller tasks.

Historical Trends

Prior cached runs (Aug-Sep 2026) showed ~84% merge rates (e.g. 0.843, 0.841, 0.839). Today's 83.6% is consistent, with no notable shift.

---

Generated by Copilot PR Prompt Analysis (Run: 40398341943)

Generated by 🔍 Copilot PR Prompt Pattern Analysis · copilot · auto · ~14 AIC · ⌖ ~1 AIC · ⊞ 11.6K · ◷

• expires on Oct 27, 2026, 12:50 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 #72513.