skill-quality-reviewer
Pass
Audited by Gen Agent Trust Hub on Jul 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local shell and Python scripts (
scripts/extract-yaml.shandscripts/skill-audit.py) to process target skill directories. While these are internal scripts, they accept file paths as arguments which could potentially lead to command injection if paths are not strictly validated by the agent before execution. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8). Its core function involves reading and analyzing the content of other skills' documentation (
SKILL.md), which is considered untrusted data. A malicious skill being reviewed could contain hidden instructions designed to manipulate the auditing agent's output or behavior. - Ingestion points: Reads the content of
SKILL.mdand referenced files from a directory provided by the user (Step 1 and Step 3 of the workflow). - Boundary markers: None. The skill does not use specific delimiters or instructions to ignore embedded prompts when parsing the target skill's content.
- Capability inventory: The agent has the capability to read local files, execute local scripts, and write new markdown files to the local directory.
- Sanitization: There is no evidence of sanitization or filtering of the content read from the target skills before it is processed for scoring and report generation.
Audit Metadata