[prompt-analysis] Copilot PR Prompt Analysis - 2026-10-02

[github-actions[bot]]

🤖 Copilot PR Prompt Pattern Analysis - 2026-10-02

Summary

Analysis Period: Last 30 days
Total PRs: 592 | Merged: 454 (76.7%) | Closed: 135 (22.8%) | Open: 3

Categories are assigned by title keyword matching (first match wins). All PRs had non-empty bodies.

Prompt Categories and Success Rates

Category	Total	Merged	Success Rate (of completed)
Bug Fix	106	93	87.7%
Feature Addition	119 (2 open)	74	63.2%
Refactoring	13	11	84.6%
Documentation	20	14	70.0%
Testing	14	10	71.4%
Other	320 (1 open)	252	79.0%

Prompt Analysis

✅ Successful Prompt Patterns

• Average body length: 183.8 words (merged)
• Bug-fix/refactor framing with a narrow target merges most often (85–88%).
• 128 of 454 merged PRs (28%) mention file paths in the body.

Examples:

1. PR Embed the bootstrap registration template safely #64974: Embed the bootstrap registration template safely → Merged
2. PR Document SEC-005 boundary for mention resolution #64973: Document SEC-005 boundary for mention resolution → Merged
3. PR Claude: default to strict MCP configuration #64970: Claude: default to strict MCP configuration → Merged

❌ Unsuccessful Prompt Patterns

• Average body length: 345.2 words (nearly 2× merged)
• Feature-titled PRs close at ~37%.
• Broad/structural or placeholder tasks dominate closures.

Examples:

1. PR Implementing feature related to ADR-64955 #65028: Implementing feature related to ADR-64955 → Closed
2. PR Route model classification from workflow task text #64859: Route model classification from workflow task text → Closed
3. PR Split workflow resolution into focused modules #64854: Split workflow resolution into focused modules → Closed
4. PR Dispatch Work Coordinator — implementation pending #64853: Dispatch Work Coordinator — implementation pending → Closed

Key Insights

• Length: Closed PR bodies are ~88% longer than merged ones, suggesting oversized scope correlates with rejection.
• Type: Features have the lowest merge rate (63%) vs bug fixes (88%).
• File references: Merged PRs mention files in 28% of bodies vs 58% of closed — bodies listing many files likely reflect broad scope rather than helping.

Recommendations

1. DO: Keep tasks narrowly scoped, one defect/behavior per prompt.
2. DO: Split large features into small incremental prompts with explicit acceptance criteria.
3. AVOID: Vague or placeholder prompts ("implementation pending", "feature related to ADR-x") and multi-module restructurings.

Historical Trends

Date	PRs	Merge Rate	Top Category
2026-10-02	592	77.1%	Other (320)
2026-10-01	566	77.4%	Other (274)

Trend: Stable merge rate (~77%); closed-PR length remains ~2× merged length.

---

Generated by Copilot PR Prompt Analysis (Run: 37062404167)

Generated by 🔍 Copilot PR Prompt Pattern Analysis · copilot · auto · 14.1 AIC · ⌖ 6.58 AIC · ⊞ 10.8K · ◷

• expires on Oct 3, 2026, 12:49 PM UTC-08:00

[github-actions[bot]]

This discussion was automatically closed because it expired on 2026-10-03T20:49:09.007Z.

Closed by Workflow