Claude Report — 2026-04-27
- Anthropic publishes election safeguards update ahead of 2026 US midterms; Opus 4.7 passes 100% of safety evals.
- Rate Limits API ships April 24, letting admins programmatically query org and workspace rate limit configs.
- `claude-agent-sdk-python` v0.1.67 fixes a Trio async compatibility regression via sniffio-based dispatch.
- yurukusa documents six weeks of Claude Code disruptions in a paid "Migration Playbook" preview gist.
- Claude Code system prompt published as a gist, accrues 173 stars in under 24 hours.
🚀 Anthropic Official
Election safeguards update for 2026 midterms
Anthropic documents its pre-election measures: Claude cannot be used for deceptive political campaigns, voter fraud, or election interference; automated classifiers and a dedicated threat intelligence team enforce the policy. In red-team tests against 600 paired prompts, Opus 4.7 responded appropriately 100% of the time and Sonnet 4.6 99.8%; both models scored 95–96% on political impartiality evaluations; new TurboVote banners now surface in-product for US and Brazil voting queries (Apr 24).
Rate Limits API — GA
Launches programmatic read access to the rate limits configured for an organization and its workspaces, complementing the existing Usage & Cost API and the rate limit charts in the Console; aimed at admins who need to monitor headroom or audit workspace quotas without screen-scraping the dashboard (Apr 24).
🔌 Claude Code Plugins
None of the top 10 updated in window.
Unchanged in window: frontend-design, superpowers, context7, code-review, code-simplifier, github, skill-creator, playwright, feature-dev, claude-md-management
🛠️ Skills
None of the top 5 updated in window.
Unchanged in window: superpowers:brainstorming, superpowers:tdd, superpowers:systematic-debugging, superpowers:writing-plans, superpowers:executing-plans
🤖 Agents & Subagents
claude-agent-sdk-python v0.1.67 / v0.1.68
v0.1.67 (Apr 25, 00:27 UTC) fixes a RuntimeError: no running event loop crash when using ClaudeSDKClient or query() under Trio — uses sniffio-based dispatch to select the correct async primitive at runtime and adds sniffio>=1.0.0 as an explicit dependency; the bundled Claude CLI was briefly v2.1.120. v0.1.68 (Apr 25, 02:07 UTC) immediately rolls the bundled CLI back to v2.1.119, confirming the v2.1.120 revert chain is complete.
Unchanged in window: superpowers:subagent-driven-development, claude-router:orchestrate, ralph-loop, claude-router:fast-executor, claude-mem:do
🔗 MCPs & Integrations
None of the top 5 updated in window.
Unchanged in window: modelcontextprotocol/servers, github-mcp-server, filesystem MCP, claude-mem, context7
💡 Community — Workflows & Ideas
Claude Code six-week disruption timeline, March–April 2026
yurukusa publishes a dated, sourced chronology of compounding Claude Code incidents — cache TTL reductions, Opus 4.7 cost inflation, quota-drain reports, and the Pro-plan removal reversal — framed as a preview chapter from a paid Migration Playbook ($19, shipping May 8) to help teams decide whether to stay, migrate, or go hybrid (Apr 24).
Claude Code full system prompt published
A gist containing Claude Code's current system prompt gathers 173 stars and 70 forks within 24 hours of publication; note that the actual prompt is dynamically generated per session, so this captures one concrete snapshot rather than a canonical definition (Apr 26).
📰 Quick Mentions
- quemsah/awesome-claude-plugins — automated n8n-based tracker of Claude plugin adoption metrics across GitHub hit 14,023 indexed repos (Apr 26).
- Releasebot Claude Code log — confirms no new Claude Code version shipped Apr 25–27; v2.1.119 remains current after the v2.1.120 auto-revert.