Unverified50% confidenceEventExact time
Anthropic's first patch blocked fetch() and XMLHttpRequest calls to file:// within the page, but Noma successfully bypassed it by moving fetch() into a Web Worker created from a blob: URL.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/18/2026
First Seen
Valid until: 1/16/2027
Sources
Related Entities
Related Claims
UnverifiedAnthropic's server-side tools (such as web search and computer use) and PDF file parsing are unavailable under this scheme because they run on Anthropic's side; sending these non-standard fields typically returns HTTP 400 (Bad Request).63% similarUnverifiedThe Fetch Web Page node is essentially an HTTP request tool that accesses the website URL registered by a business and determines whether the site is accessible based on the returned HTTP status code.61% similarUnverifiedThe behavior of URL fragments not being sent to the server is established by the RFC 3986 specification, originally designed for in-page navigation.61% similarUnverifiedUsers of third-party Anthropic-compatible endpoints (ANTHROPIC_BASE_URL) have been receiving HTTP 400 on every turn since 2.1.265; the root cause was a regular expression in the Artifact tool input schema being rejected by these endpoints, and the new version has fixed this.60% similarUnverifiedThe complete crawler workflow is: initiate HTTP request → retrieve HTML source code → parse and extract → store data59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1061014API
curl https://kongchang.com/api/v1/knowledge/claims/1061014MCP
get_claim(id=1061014)