The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to install the crlfuzz tool using go install github.com/dwisiswant0/crlfuzz/cmd/crlfuzz@latest. This fetches code from a third-party GitHub repository that is not managed by a verified or trusted vendor.
[COMMAND_EXECUTION]: The documentation includes multiple shell commands for tool installation and execution, including piping data from other tools (cat urls.txt | crlfuzz).
[PRIVILEGE_ESCALATION]: The installation instructions include the use of sudo apt install crlfuzz, which requires administrative privileges to modify the system's package state.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to process external lists of URLs that could contain malicious payloads or instructions.
Ingestion points: The skill reads input from external files (-l urls.txt) or piped standard input.
Boundary markers: There are no instructions or delimiters defined to prevent the agent from being influenced by data contained within the processed URL lists.
Capability inventory: The skill leverages shell execution, network scanning capabilities (via crlfuzz), and file system writes for logging results.
Sanitization: The instructions do not specify any validation or filtering of the input URLs before they are passed to the scanning tool.

crlfuzz