The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The research-summarizer sub-skill includes instructions to clone a repository from github.com/alirezarezvani. This organization is not on the trusted vendor list, and downloading code from unknown third-party repositories for local execution poses a risk of script compromise.
[INDIRECT_PROMPT_INJECTION]: The toolkit contains multiple specialized tools that ingest and analyze external, untrusted content, creating a risk that instructions embedded in that data could influence the agent's behavior.
Ingestion points: Analysis of arbitrary codebases in code-to-prd, processing of external research papers and web articles in research-summarizer, and synthesis of customer interview transcripts in product-manager-toolkit.
Boundary markers: Absent. The instructions do not specify the use of delimiters (like XML tags or triple quotes) or provide guidance to the agent to ignore instructions found within the processed data.
Capability inventory: The skill uses Python scripts to perform deep analysis and can write complex file structures (PRDs, codebases, boilerplates) to the local filesystem.
Sanitization: No sanitization or filtering logic is mentioned in the provided instructions for handling the external text inputs.

product-skills