The Agent Skills Directory

[DATA_EXFILTRATION]: Accesses sensitive configuration files and user home directory paths to audit for credential exposure and infection markers.
The scripts/scan-npm.sh scanner inspects ~/.npmrc, PROJECT_DIR/.npmrc, and /etc/npmrc for _authToken strings.
The scripts/scan-python.sh scanner checks ~/.pypirc for password fields and API tokens.
The scanners check various paths in the user's home directory (e.g., ~/.local/share/pgmon/service.py, ~/.config/sysmon/sysmon.py) for known malware persistence indicators.
[COMMAND_EXECUTION]: Relies on the execution of multiple bash scripts provided within the skill to perform environmental and project auditing.
Orchestrates scanning through scan-all.sh, which calls individual scanners for npm, Python, and CI/CD pipelines.
Uses standard utilities such as grep, find, systemctl, and crontab to inspect the local system state and project files.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (project dependencies, lockfiles, and configuration) which creates a surface for indirect instruction injection.
Ingestion points: Multiple scanners (scan-npm.sh, scan-python.sh, scan-ci.sh) read project files like package.json, requirements.txt, and GitHub Actions workflows.
Boundary markers: Lacks explicit delimiters or instructions to ignore embedded commands when processing these files.
Capability inventory: Includes shell execution, file reading, and tool invocation (pip, python3).
Sanitization: No explicit sanitization of project file contents is performed prior to analysis.
[DYNAMIC_EXECUTION]: Executes Python code at runtime to identify system-specific paths for auditing.
Uses python3 -c "import site; ..." to dynamically locate site-package directories during the scanning process.

supply-chain-guard