price-target-consensus
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill setup instructions use
npxto fetch and run theoctagon-mcppackage. This is a vendor-owned resource from OctagonAI required for the skill to communicate with the financial data server. - [EXTERNAL_DOWNLOADS]: The documentation references the official Homebrew installation script from a well-known repository (
github.com/Homebrew). This is a standard prerequisite for Mac users to install Node.js. - [REMOTE_CODE_EXECUTION]: The prerequisite setup includes a piped shell command to install Homebrew (
curl ... | bash). This targets the official Homebrew installation source and is considered standard practice for infrastructure setup. - [INDIRECT_PROMPT_INJECTION]: The skill ingests external data by calling the
octagon-agenttool to retrieve stock metrics. This represents a potential injection surface, though no exploitable capabilities are present within the skill to leverage this surface. - Ingestion points: Financial target data returned from the
octagon-agenttool referenced inSKILL.md. - Boundary markers: None used in the prompt argument for the tool call.
- Capability inventory: The skill does not define any file-write, network exfiltration, or subprocess execution capabilities that could be misused.
- Sanitization: None observed in the processing of tool output.
Audit Metadata