Claude Code v2.1.280 adds Opus 5.5 with 1M context at $4/$20 per Mtok
ashwin-ant · ghdev · 2026-09-23
anthropics/claude-code shipped v2.1.280:
- Adds Claude Opus 5.5 (claude-opus-5-5) as the default Opus model: 1M context, $4/$20 per Mtok with $0.20/Mtok cache reads
- Mouse support extended to more fullscreen lists (/skills scrolling, /plugin skill state options)
- New CLAUDECODEMAXMCPDESCRIPTIONLENGTH to adjust the 2,048-char MCP tool description cap
- Security fixes: writes through symlinked paths are judged by where they land (acceptEdits, allow rules and auto mode no longer approve out-of-tree writes); auto mode stops endlessly retrying actions declined by safety checks and backs off when checks return no answer; validation fixes for Write calls with wrong field names
- UX fixes: double Ctrl+C/D no longer quits, stray n/y no longer close or confirm dialogs, first-click-to-focus no longer triggers the item underneath
More from coding & agent
- Security audit: autonomous research program XBOW credited with ~12 upstream bug fixes — moyix · 2026-09-23
- Telling agents to use formal verification helps them write better code — sh_reya · 2026-09-23
- Team uses Typesafe's Jev to prompt follow-up questions during ticket creation — TheMoonMidas · 2026-09-23
- The most common evals mistake: skipping error discovery and measuring the wrong thing — FinanceYF5 · 2026-09-23
- AI products are easy to change and hard to predict: evals turn 'good' into repeatable tests — FinanceYF5 · 2026-09-23
- Is MCP a Bad Idea? New Video Questions the Agent Protocol Standard — RelevantEmergency707 · 2026-09-23