The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches metadata, repository descriptions, and README files from the official GitHub API (api.github.com) to identify red flags and malicious patterns. These network operations target a well-known and trusted service for legitimate analysis purposes.
[COMMAND_EXECUTION]: Provides illustrative shell scripts and Python code snippets that interact with system tools like curl, jq, and the GitHub CLI (gh). These are used to automate the inspection of repository data and to facilitate reporting of malicious content to platform administrators.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of repository descriptions and names fetched via the GitHub API. While this creates a theoretical ingestion surface for adversarial instructions, the skill handles this data through structured analysis (keyword filtering, metrics checking, and string normalization) to minimize risk. This behavior is necessary for its primary function of security auditing.
[SAFE]: No instances of obfuscation, data exfiltration, or unauthorized privilege escalation were detected. The skill promotes security best practices, such as verifying file integrity using SHA-256 hashes and obtaining software only from official vendor websites.

identify-malicious-repository