This release includes 1 security fix for security teams reviewing exposed deployments.
Topics
+7 more
Affected surfaces
Summary
AI summaryRemoved leaked files from sdist containing live API keys and internal documents.
Full changelog
Security
- Removed leaked files from sdist: Previous versions (7.6.1–8.0.3) shipped
.env(live API keys),.internal/(strategy docs),CLAUDE.md, and other internal files in the PyPI source distribution. All affected keys have been rotated. - Comprehensive
[tool.hatch.build.targets.sdist] excludelist now only ships:src/,pyproject.toml,LICENSE,README.md,CHANGELOG.md,CONTRIBUTING.md
Fixed
- Session savings report audit: 8 fixes to the session-end summary for trust and accuracy
- Subscription deltas: show "no change" instead of misleading
+0.0pp - Mock/test models filtered from production reports
- Call reconciliation line:
Total: N routed = X local/prepaid + Y external - Savings transparency: show actual + baseline cost
- "Free models" → "Local models (Ollama)" / "Prepaid models (Codex)"
- Router efficiency: honest "No fallbacks (N decisions)" instead of vague "100% on-target"
- Subscription deltas: show "no change" instead of misleading
- Star CTA SVG: redesigned layout, no text/icon overlap
- Stars badge: cache-bust to clear stale "invalid" state
Tests
- 30 new tests for session report formatting and data integrity
- 1768 total tests passing
Security Fixes
- Removed `.env` with live API keys, `.internal/`, `CLAUDE.md`, and other internal files from sdist in versions 7.6.1–8.0.3; all affected keys have been rotated.
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About ypollak2/llm-router
Subscription-aware LLM router for Claude Code. Routes tasks to 20+ providers (OpenAI, Gemini, Groq, Ollama, Codex) based on complexity classification, Claude subscription pressure, and cost. Free tasks stay on Claude subscription; expensive tasks fall back to the cheapest capable model. Includes 30 MCP tools, 6 auto-routing hooks, semantic dedup cache, prompt caching, daily spend cap, and a live web dashboard.
Related context
Related tools
Beta — feedback welcome: [email protected]