Anthropic’s Claude Code Update Sparks Developer Concerns Over Hidden File Actions
Anthropic, an AI safety and research company, recently released an update to its Claude Code platform – version 2.1.20 – that is now drawing criticism from developers. The update reportedly masks file actions performed by the AI, leading to concerns about transparency and control.
The core of the issue centers around Claude Code’s handling of files. According to reports, the latest version obscures the specific actions taken on files, displaying only a general indication that a file has been modified rather than detailing the changes made. This lack of granularity has prompted a negative reaction from some developers who rely on precise tracking of file manipulations for debugging and version control.
Anthropic is focused on building reliable, interpretable, and steerable AI systems. The company released Claude Opus 4.6 on February 5, 2026, described as its most capable model to date, bringing novel levels of reliability and precision to coding, agents, and enterprise workflows. The update signals a growing need for developers to understand and trust the inner workings of AI-powered coding tools.
Claude Opus 4.6 features a 1M context window and is available on Claude for Pro, Max, Team, and Enterprise users, as well as through the Claude Developer Platform, Amazon Bedrock, Google Cloud’s Vertex AI, and Microsoft Foundry. Pricing starts at $5 per million input tokens and $25 per million output tokens.
Anthropic also recently released Claude Sonnet 4.6 on February 17, 2026, offering frontier performance across coding, agents, and professional function.
The company provides resources for developers through Anthropic Academy, offering API development guides and enterprise deployment best practices. Featured courses include those focused on AI Fluency, API development, Model Context Protocol and Claude Code.
