Code Smells in Clojure: Initial Findings from a Grey Literature Review
2026-05-26T08:51:49Z•4e933d4cc8220fcd5dc068523008abb262df2b15a0a6a8aec278edc6ee09376e
BugsRepoClojureLLMsTRACERbreaking changesbug reportscode modelscode smellsdata contaminationdependency managementevaluation harnessesgraph neural networkshallucination detectionmicroservice intrusion detectionmulti-agent programmingprompt pattern frameworksecure coding educationskill librariesskill shadowingsoftware visualization
What happened
Collection of recent software-engineering and ML-for-code research from arXiv covering: (1) a grey-literature catalog of 26 Clojure-specific and functional code smells and gap in linter support; (2) analysis of LLM agent skill libraries showing performance degradation driven by "skill shadowing" as libraries scale; (3) TRACER, a semantic-aware fine-grained code-contamination detection framework and benchmark (functionally identical / nearly identical / shared logic); (4) section-aware hallucination detection for LLM-generated bug-report summaries using a BugsRepo-derived benchmark; (5) an HCI/
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- arxiv_cs_se
- Record identifier
- 4e933d4cc8220fcd5dc068523008abb262df2b15a0a6a8aec278edc6ee09376e
- Enrichment time
- 2026-05-26T08:51:49Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.