skills/agentskillexchange/skills/Run Claude Code job-search operations with Career-Ops/Gen Agent Trust Hub
Run Claude Code job-search operations with Career-Ops
Warn
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: MEDIUMMETADATA_POISONINGINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [METADATA_POISONING]: The skill metadata explicitly claims 54,286 GitHub stars across multiple fields. This number is statistically improbable for a niche CLI tool and is considered deceptive metadata intended to induce unearned trust in the skill's safety.
- [INDIRECT_PROMPT_INJECTION]: The skill's workflow is centered on ingesting external job listings, which creates a significant vulnerability where malicious instructions could be embedded in listing data to manipulate agent actions.
- Ingestion points: Job descriptions and external web content fetched via browser automation tools (SKILL.md).
- Boundary markers: No delimiters or safety instructions are defined to separate untrusted data from the agent's core prompt.
- Capability inventory: Uses Playwright for browser automation and Claude Code for content analysis and generation (SKILL.md).
- Sanitization: No validation or sanitization procedures for ingested listing data are described.
- [NO_CODE]: The skill body contains no executable instructions, scripts, or operational prompts, serving purely as a metadata container for an external repository.
Audit Metadata