mirror of
https://github.com/0xMassi/webclaw.git
synced 2026-06-09 22:35:12 +02:00
When bash splits a URL at & or ? (a common foot-gun), webclaw
receives only the truncated prefix and silently fetches the wrong
page. Per issue #6:
1. Heuristic warning: if the URL ends with '&' or contains '?' with
no '=' after, emit a stderr warning before fetching:
# webclaw: warning: URL looks truncated (ends with '&' or '?'); did the shell split it? Quote the URL or use --url-encoded.
2. New flag --url-encoded: parallel input that asserts the user has
handled escaping. Suppresses the truncation warning since intent
is explicit.
Fetch proceeds in both cases; this is informational only. 4 new
tests in webclaw-cli. Workspace 720 -> 724.
(cherry picked from commit
|
||
|---|---|---|
| .. | ||
| webclaw-cli | ||
| webclaw-core | ||
| webclaw-fetch | ||
| webclaw-llm | ||
| webclaw-mcp | ||
| webclaw-pdf | ||
| webclaw-server | ||