The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill uses the powershell -ExecutionPolicy Bypass flag when executing the render_resume.ps1 script. This flag is designed to override system security settings that restrict script execution, effectively bypassing the local security policy.
[COMMAND_EXECUTION]: The skill facilitates the execution of various shell commands, including PowerShell, Python, and npx, to manage resume rendering, validation, and editing tasks as part of its core workflow.
[EXTERNAL_DOWNLOADS]: The skill uses npx to download and execute the @chasen-liao/resume-skills package from the NPM registry at runtime. While this package is associated with the skill vendor, fetching the latest version from an external registry introduces a dependency on remote code.
[INDIRECT_PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its ingestion of untrusted data from user-provided resumes and job descriptions. Evidence chain: 1) Ingestion points include PDF, DOCX, and HTML resumes as well as JD text (SKILL.md). 2) Boundary markers are partially present through instructions to verify parsed facts before inclusion. 3) Capability inventory includes shell execution (PowerShell, Python, npx) and file system writes. 4) No explicit sanitization or filtering of the ingested content is documented.

resume-workflow