pr-launch-pack
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple external sources.
- Ingestion points: Content is ingested from fetched web pages, uploaded files, and user-pasted text for claim verification (SKILL.md, Step 2).
- Boundary markers: Explicit instructions are provided to treat third-party content as data only and to ignore any embedded instructions (SKILL.md, Introduction and Step 2).
- Capability inventory: The skill utilizes a local Python script (scripts/text_check.py) and possesses web search/fetch capabilities.
- Sanitization: The skill includes a specific rule to detect and report likely scams if personal, tax, or bank details are found in ingested data (SKILL.md, Step 2).
- [COMMAND_EXECUTION]: The skill executes a local Python script (scripts/text_check.py) to perform character counting, date verification, and dash detection. The script is part of the skill package and uses standard Python libraries without performing network operations or unsafe dynamic code execution.
Audit Metadata