The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references several external Python packages and GitHub repositories common in the CTF community (e.g., pycryptodome, z3-solver, RsaCtfTool). These are standard tools for cryptographic analysis and do not point to suspicious or untrusted sources.
[INDIRECT_PROMPT_INJECTION]: The skill provides numerous command-line and code templates where user-supplied data (such as ciphertexts or RSA parameters) is intended to be interpolated. This creates a potential surface for indirect prompt injection if the agent does not properly sanitize these inputs before execution.
[SAFE]: No malicious patterns such as credential exfiltration, persistence mechanisms, or unauthorized privilege escalation were detected. The skill's behavior is entirely consistent with its stated cryptographic research and competition focus.

ctf-crypto