positioning
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data regarding product names, buyer context, and competitive alternatives in its workflow sections. It lacks explicit boundary markers or instructions for the agent to ignore commands potentially embedded within these user inputs. Ingestion points include the 'Define the positioning situation' and 'Identify true competitive alternatives' tables. Capability inventory: The skill utilizes platform-level 'Read' and 'Grep' tools but contains no internal scripts for command execution or file modification. Sanitization: No sanitization or validation of user-provided content is mentioned.
- [EXTERNAL_DOWNLOADS]: The skill's metadata contains informational references to external domains (aprildunford.com, public.summaries.com) for grounding the methodology in universal truth sources. These are informational references and do not trigger automatic execution.
- [NO_CODE]: The skill consists entirely of instructional markdown and YAML configuration, lacking any scripts, binaries, or executable components, which minimizes the direct attack surface.
Audit Metadata