The Agent Skills Directory

[SAFE]: The skill provides standard reconnaissance and analysis command snippets (e.g., file, strings, binwalk, checksec, nc) for security researchers to analyze challenge files and connect to remote services in a controlled environment.- [PROMPT_INJECTION]: The skill specifies workflows for analyzing untrusted external artifacts, which creates a surface for indirect prompt injection.
Ingestion points: External CTF challenge files processed through utilities such as file, strings, binwalk, and xxd as defined in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore potentially malicious embedded instructions in the challenge data.
Capability inventory: The skill possesses capabilities for local file system analysis and network connectivity using the nc command.
Sanitization: No sanitization or validation of content extracted from processed files is described in the provided instructions.

ctf-solve