First Seen
Feb 18, 2026
Last Scanned
Feb 22, 2026
Findings
3
Score
100/100
Findings (3)
Detects patterns where external API responses are used directly without validation or sanitization
API + data + directly applyValidate and sanitize all data received from external APIs before using it in tool operations or agent prompts. Implement schema validation and treat API responses as untrusted input.
Likely FP if the match is a truncated table cell or documentation fragment that mentions API responses in a descriptive context, not actual unvalidated data processing.
Detects tmux/screen send-keys used to inject commands into terminal sessions
tmux send-keys Avoid sending unsanitized user input to tmux/screen sessions via send-keys. Use a controlled command dispatch mechanism instead of injecting commands into terminal multiplexers.
Likely FP if the match is in documentation describing tmux/screen workflow automation for the user themselves, not controlled by an external agent.
Detects tmux/screen send-keys used to inject commands into terminal sessions
tmux send-keys Avoid sending unsanitized user input to tmux/screen sessions via send-keys. Use a controlled command dispatch mechanism instead of injecting commands into terminal multiplexers.
Likely FP if the match is in documentation describing tmux/screen workflow automation for the user themselves, not controlled by an external agent.