The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill attempts to harvest GitHub authentication tokens from sensitive local files including ~/.hermes/.env and ~/.git-credentials to facilitate API authentication.
[COMMAND_EXECUTION]: The skill uses python3 -c to execute inline Python scripts for parsing JSON data received from GitHub API calls. This is used for extracting specific fields from API responses.
[REMOTE_CODE_EXECUTION]: Automated scanning flagged a pattern where curl output is piped to python3. Detailed analysis confirms this is used for parsing structured JSON data from the official GitHub API using static Python code provided via the -c flag, which does not constitute execution of arbitrary remote scripts.
[EXTERNAL_DOWNLOADS]: The skill performs repository cloning and downloads release assets from GitHub. These operations target well-known GitHub domains.
[PROMPT_INJECTION]: The skill processes untrusted data from the GitHub API (such as repository descriptions and issue titles) which could contain malicious instructions. This represents a potential indirect prompt injection surface.
Ingestion points: Data from api.github.com endpoints, including repository details, search results, releases, and workflow logs.
Boundary markers: Absent; external content is printed directly to the output without delimiters or instructions to ignore embedded commands.
Capability inventory: Access to git, gh CLI, and shell command execution through the agent environment.
Sanitization: No sanitization or validation of the remote content is performed before it is processed or displayed.

github-repo-management