Anthropic debuts Claude Code Assist Pro with 2M-token context for dev teams

AI · 5 min read

Anthropic released Claude Code Assist Pro, targeting engineering teams that need deep context for large codebases. The product supports a 2 million token context window, enabling the model to ingest entire repositories or long-running CI logs for end-to-end code suggestions and security checks.

Pro features include a structured code review assistant that highlights potential security regressions, license conflicts, and performance anti-patterns. Teams can also define organizational guardrails—validated prompts and suppression rules—so the assistant aligns with existing review norms.

Anthropic is offering enterprise deployment options, including VPC peering and on-premise model hosting for regulated customers. The company emphasized transparency, providing audit logs of prompts and responses to help compliance teams trace recommendation origins.