marketing-pipeline-automated-content
Warn
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to clone project source code from an unverified third-party GitHub repository (
pennydinh/marketing-pineline-share.git). - [REMOTE_CODE_EXECUTION]: The installation and usage instructions involve running
npm installand executing CLI commands from the cloned repository, which results in the execution of code from an external, untrusted source. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from uncontrolled external sources for processing by an LLM.
- Ingestion points: The
crawlNewsfunction retrieves data from TechCrunch, a16z, Twitter, and LinkedIn. - Boundary markers: The provided documentation and code examples do not use delimiters or instructions to prevent the LLM from obeying commands embedded in the crawled content.
- Capability inventory: The skill possesses the capability to write to the file system (via
renderMediaoutputting to./output/) and perform network requests via crawlers. - Sanitization: There is no evidence of input sanitization or validation before crawled data is passed to the AI analysis functions.
- [DYNAMIC_EXECUTION]: The skill utilizes
@remotion/bundlerto programmatically bundle code at runtime and@remotion/rendererto execute that code for media generation, which can be exploited if the inputs to the bundling process are compromised.
Audit Metadata