Beyond "What to Retrieve": Uncertainty in Retrieval-Augmented Code Generation

arXiv 2607.24884•54a95db5012a75c0b951f9b5cf85004dd96c07565f1431dc53903cae89f07488
AI-assisted codingDevOpsDockerJavaScriptLLMagent skillsautomated bug findingcode generationcontainer securitydeployment securitydeveloper editsformal verificationnetwork segmentationrepository contextretrieval-augmented generationsecrets managementsecurity requirementssoftware engineeringsupply-chain securitytraceability

Paper metadata

arXiv ID
2607.24884
Version
Not specified by this published record
Category
Computer Science — Software Engineering (cs.SE)

The PDF link points to arxiv.org. Baitaphish does not expose a private stored PDF.

Evidence and limitations

Source ID
arxiv_cs_se
Record identifier
54a95db5012a75c0b951f9b5cf85004dd96c07565f1431dc53903cae89f07488
Enrichment time
2026-07-29T08:51:39Z
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.

Beyond "What to Retrieve": Uncertainty in Retrieval-Augmented Code Generation · Baitaphish