Fetch logs with optional filtering
Fetch log lines newest-first, paging backwards towards the start of the file. When level or keyword is set the server keeps reading backwards until the requested number of matching lines has been collected, the start of the file is reached, or the per-request scan cap (32 MiB) is hit — so has_more means "more matches may exist", not merely "more bytes exist". Offsets in next_cursor are byte offsets into the file the page was read from; for a .log.gz archive they address the uncompressed stream.
Query parameters
Number of log lines to fetch (default is 100). With level or keyword set this is the number of matching lines.
Filter logs by level (e.g., INFO, ERROR, DEBUG). Matched as a substring, searched backwards across the whole file rather than within one page.
Filter logs containing a specific keyword or phrase. Matched as a substring, searched backwards across the whole file rather than within one page.
Opaque pagination cursor from a previous response's next_cursor; fetches the next page.
Specific log file to read (default is the current log file). Rotated .log.gz archives are accepted and decompressed transparently.
Response
Logs fetched successfully