employee-engagement

Warn

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to automatically execute a shell command to probe for installed skills in multiple hidden directories such as ~/.claude/skills/, ~/.agent/skills/, and ~/.agents/skills/ during the first activation. This constitutes local environment fingerprinting without user initiation.
  • [EXTERNAL_DOWNLOADS]: The skill instructions recommend downloading and executing code from an external repository using npx skills add AbsolutelySkilled/AbsolutelySkilled. This repository is not identified as a trusted source, introducing a supply chain risk.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it is designed to ingest and analyze external, potentially attacker-controlled survey data while having the capability to execute shell commands.
  • Ingestion points: Employee survey results, pulse check data, and qualitative open-text responses processed in SKILL.md.
  • Boundary markers: Absent; the skill lacks delimiters or explicit instructions to treat survey data as untrusted data.
  • Capability inventory: Command execution (ls) and the ability to recommend further installation of external code (npx).
  • Sanitization: Absent; the skill does not define any validation or filtering for the survey data it processes.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 7, 2026, 03:04 PM
Security Audit — agent-trust-hub — employee-engagement