wilco-ad-setup
Pass
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to execute shell commands for environment management and data validation.
- Evidence: Use of
mkdir -p ./campaignsfor output directory creation in SKILL.md. - Evidence: Use of
awk -F',' '{print NF}' ...for CSV column count validation in the validation section of SKILL.md. - [DYNAMIC_EXECUTION]: The skill utilizes Python script generation and execution to handle complex CSV formatting tasks.
- Evidence: SKILL.md provides a Python template using the
csvmodule to generate the final campaign Bulksheet file. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from Amazon research tools, creating a vulnerability surface where malicious instructions in keyword data could influence agent behavior.
- Ingestion points: Keyword lists and competitor ASIN data ingested from
research_productsandamazon_keyword_researchin Phase 2 of SKILL.md. - Boundary markers: Absent; the sub-agent prompt template in Phase 2 lacks delimiters or specific instructions to ignore embedded commands within the keyword strings.
- Capability inventory: The skill can write files to the local system and execute shell/Python commands.
- Sanitization: Absent; while the skill filters for relevance and search volume, it does not sanitize retrieved strings for prompt injection patterns.
Audit Metadata