This release adds 2 notable features for engineering teams evaluating rollout.
✓ No known CVEs patched in this version
Summary
AI summaryUpdates Verification, Highlights, and Upstream attribution across a mixed release.
Changes in this release
| Type | Severity | Summary | CVE |
|---|---|---|---|
| Feature | Medium |
Adds configurable extraction-provider priority via `auto_routing.extract_provider_priority` or CLI command. Adds configurable extraction-provider priority via `auto_routing.extract_provider_priority` or CLI command. Source: llm_adapter@2026-07-15 Confidence: high |
— |
| Feature | Medium |
Increases default full‑content extraction budgets to 60k characters per result and 120k total. Increases default full‑content extraction budgets to 60k characters per result and 120k total. Source: llm_adapter@2026-07-15 Confidence: high |
— |
| Bugfix | Medium |
Improves cache ownership safety; cache stats and clear now manage only complete Web Search Plus search‑cache envelopes, preserving unrelated or corrupt data. Improves cache ownership safety; cache stats and clear now manage only complete Web Search Plus search‑cache envelopes, preserving unrelated or corrupt data. Source: llm_adapter@2026-07-15 Confidence: high |
— |
| Refactor | Low |
Maintains stable MCP contract exposing only `web_search` and `web_extract` with unchanged provider enums. Maintains stable MCP contract exposing only `web_search` and `web_extract` with unchanged provider enums. Source: llm_adapter@2026-07-15 Confidence: high |
— |
Full changelog
web-search-plus-mcp v0.17.0
This release synchronizes the MCP package with Web Search Plus v2.9.1 plus the post-release configurable extraction-priority feature.
Highlights
- Safer cache ownership: cache stats and clear now manage only complete Web Search Plus search-cache envelopes. Provider health/stats, host usage state, unrelated or corrupt JSON, and invalid UTF-8 files are preserved.
- Independent extraction routing: configure
auto_routing.extract_provider_priorityor useweb-search-plus-mcp config set-extract-priority ...without changing search routing. - Fairer Parallel extraction: default full-content budgets increase to 60k characters per result and 120k total; explicit config overrides still win.
- Stable MCP contract: the server continues to expose only
web_searchandweb_extract, with the same provider enums and Tavily-first public extraction default when no override is configured.
Upstream attribution
- Web Search Plus #86 — Parallel full-content budget
- Web Search Plus #88 / v2.9.1 maintenance — cache ownership safety
- Web Search Plus #89 — configurable extraction priority
Verification
- 115 tests
- Ruff and compileall
- wheel and source distribution with Twine checks
- MCP schema and stdio handshake
- live Serper search, Serper extraction, Parallel extraction, and Serper-first auto-extraction
Install or upgrade:
pip install --upgrade web-search-plus-mcp==0.17.0
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 robbyczgw-cla/web-search-plus-mcp
Multi-provider web search with intelligent auto-routing (Serper, Tavily, Exa)
Related context
Related tools
Earlier breaking changes
- v1.0.0 Native Perplexity and Kilo Perplexity answer endpoints removed from public provider schemas.
- v1.0.0 Public search surface reduced to 12 verified source-result providers; extraction limited to 8 providers.
- v0.5.1 kilo_perplexity renamed to kilo-perplexity without aliasing
- v0.5.1 Perplexity provider separated from kilo-perplexity with distinct API keys
Beta — feedback welcome: [email protected]