Steven Gonsalvez

Software Engineer

v2.1.211

Why CEREBRO kept it

Subagent text forwarding; core LLM mechanics

The text below is an automated extraction of the article at https://github.com/anthropics/claude-code/releases/tag/v2.1.211, stored verbatim in the public cerebro-vault repository. Copyright remains with the original publisher (github.com).

<h2>What's changed</h2> <ul> <li>Added <code>--forward-subagent-text</code> flag and <code>CLAUDE_CODE_FORWARD_SUBAGENT_TEXT</code> environment variable to include subagent text and thinking in stream-json output</li> <li>Fixed permission previews relayed to chat channels not neutralizing bidirectional-override, zero-width, and look-alike quote characters, so tool inputs cannot visually alter the approval message</li> <li>Fixed auto mode overriding a PreToolUse hook's <code>ask</code> decision for unsandboxed Bash — a hook <code>ask</code> now floors the decision at a prompt</li> <li>Fixed par

Backlinks

Appeared in 6 briefings

6 briefings re-surfaced this signal. A high number here is a deduplication weakness in the pipeline, not a popularity score.

Related

Shares tags: ai/agents · ai/llm-mechanics · release-notes

Also from github.com