The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to guide users to install an external dependency using the command npx skills add rendercv/rendercv-skill. This dependency is hosted in an external repository that is not part of the trusted vendor list, representing a supply chain risk.
[REMOTE_CODE_EXECUTION]: By promoting the installation of an external skill, the skill facilitates the execution of third-party code within the agent environment. The recommended command includes the -y flag, which bypasses user confirmation prompts, thereby reducing the opportunity for manual security review during installation.
[INDIRECT_PROMPT_INJECTION]: The skill processes "Career Positioning OS" resume content which acts as an untrusted ingestion point for external data.
Ingestion points: Resume and achievement content passed into the handoff prompt in SKILL.md.
Boundary markers: While the skill includes instructions to preserve metrics and factual claims, it lacks robust delimiters or "ignore embedded instructions" warnings to prevent data from being interpreted as commands by the downstream rendercv skill.
Capability inventory: The skill facilitates the installation and use of external tools with file system and rendering capabilities.
Sanitization: There is no mention of sanitization, filtering, or validation processes for the input resume data before it is handed off to the rendering pipeline.

rendercv-resume-handoff