Fingerprinting AI Coding Agents (MSR '26 Paper)

• Entity ID: ent-20260419-9bca6bf795bf
• Type: document
• Scope: shared
• Status: active
• Aliases: arXiv 2601.17406, AI agent fingerprinting paper

Description

Peer-reviewed paper (MSR '26, January 2026) that analyzed 33,580 PRs across five major AI coding agents and reached 97.2% F1-score in agent identification. Claude Code's most distinctive fingerprint is conditional-statement structure (27.2% feature importance).

Key claims

• Fingerprinting paper reaches 97.2% F1 on agent identification

Relations

• Fingerprinting AI Coding Agents (MSR '26 Paper) --[contains]--> Conditional-Statement Fingerprint

Sources

src-20260419-f46b0a2fccd9