The Agent Skills Directory

[COMMAND_EXECUTION]: The VerifyCodeSignature function in SKILL.md uses exec.Command with fmt.Sprintf to insert the filePath variable directly into a PowerShell command string. If this skill is used to analyze a repository containing files with maliciously crafted names (e.g., using shell metacharacters like ; or &), it could lead to arbitrary command execution on the host system.
Evidence: cmd := exec.Command("powershell", "-Command", fmt.Sprintf("(Get-AuthenticodeSignature '%s').Status", filePath))
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from external repositories, which creates a vulnerability surface for indirect prompt injection attacks.
Ingestion points: Repository names, descriptions, and topics are parsed into the RepositoryAnalysis struct. The file system is traversed using filepath.Walk in VerifyRepositoryContent.
Boundary markers: The instructions do not define delimiters or specific "ignore" instructions for the data being analyzed.
Capability inventory: The skill possesses the ability to execute shell commands (exec.Command) and perform network requests (net/http).
Sanitization: While the code uses strings.ToLower for keyword matching, it lacks robust sanitization for file paths or metadata before they are used in command execution or reporting logic.

malware-detection-and-removal