ANTHROPIC QUIETLY TESTS MOVING CLAUDE CODE OFF $20 PRO, THEN REVERTS — WHILE SHIPPING A SOLID V2.1.117 UPGRADE
Anthropic briefly tested removing Claude Code from new $20 Pro signups, reversed after confusion, and shipped a meaningful v2.1.117 update. The Register spotte...
Anthropic briefly tested removing Claude Code from new $20 Pro signups, reversed after confusion, and shipped a meaningful v2.1.117 update.
The Register spotted pricing pages that removed Claude Code from Pro and swapped checkmarks for Xs, while other docs still claimed access; Anthropic’s growth lead called it a 2% new-signup test and said existing subscribers weren’t affected article. Ed Zitron documented the flip and later reversal across site and docs post, and Simon Willison summarized the resulting confusion for developers blog.
Separately, Claude Code v2.1.117 landed with helpful quality-of-life and ops wins: faster startup, persistent model selection, tougher plugin dependency handling and marketplace policy enforcement, better session resume, OTEL attributes for cost/token/effort, and native BFS/ugrep-backed search on macOS/Linux (release, summary).
Seat planning and budgets may shift if Claude Code moves up-tier; today it’s a test, tomorrow it could stick.
v2.1.117 adds sturdier telemetry and plugin governance that ops teams can actually use in production.
-
terminal
Upgrade to v2.1.117 and validate OpenTelemetry emissions (cost.usage, token.usage, api_request, effort attribute) and redaction behavior; set OTEL_LOG_TOOL_DETAILS=1 in a sandbox to verify.
-
terminal
Exercise plugin flows: install/update with missing deps, blockedMarketplaces/strictKnownMarketplaces enforcement, and concurrent MCP server connect; confirm startup latency improvements on dual local+claude.ai MCP.
Legacy codebase integration strategies...
- 01.
If you rely on Pro seats for CLI/agents, run a contingency check: can critical repos move to Max/Enterprise quickly if Pro loses Code?
- 02.
Pin claude-code versions in dev containers/CI; the new plugin marketplace policies and native search tools may affect scripts and local workflows.
Fresh architecture paradigms...
- 01.
Assume coding agents live on higher tiers; bake seat costs and API telemetry into your TCO from day one.
- 02.
Standardize on MCP for tools and require OTEL traces/metrics for agent runs to keep costs and behavior observable.
Get daily ANTHROPIC + SDLC updates.
- Practical tactics you can ship tomorrow
- Tooling, workflows, and architecture notes
- One short email each weekday