The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The README.md and SKILL.md files provide instructions to install the skill collection using npx ai-agent-skills install alirezarezvani/claude-skills/ra-qm-team. This facilitates the download and execution of code from an unverified third-party GitHub repository.
[INDIRECT_PROMPT_INJECTION]: The skill suite ingests untrusted external data and possesses significant capabilities, representing a vulnerability surface.
Ingestion points: The scripts mdr_compliance_checker.py (ingests data from product folders), vulnerability_tracker.py (ingests security-scan.json), and qms_compliance_checker.py (ingests organization profiles) read external files into the agent context.
Boundary markers: There are no instructions for using delimiters or boundary markers to prevent the agent from executing instructions potentially hidden in the ingested compliance documentation or scan results.
Capability inventory: The collection includes scripts capable of writing files (technical_file_builder.py, dpia_generator.py), managing databases (capa_tracker.py), and generating automated reports across multiple modules.
Sanitization: The documentation does not describe any validation, filtering, or sanitization of the external content before it is processed by the Python tools or interpolated into the agent's context.
[METADATA_POISONING]: There is a consistency issue in the metadata where the files (SKILL.md, plugin.json, README.md) attribute the skill to 'Alireza Rezvani', while the deployment context identifies the author as 'iwanhe'. This discrepancy could lead to misjudgment of the skill's origin and safety.

ra-qm-skills