The Agent Skills Directory

[PROMPT_INJECTION]: The skill has a potential surface for indirect prompt injection by processing data from an external source without explicit safety boundaries.
Ingestion points: Data is retrieved from the Sutra (Arketa) Partner API and stored in a local SQLite mirror accessed by the agent.
Boundary markers: There are no defined instruction delimiters or boundary markers to separate data content from agent instructions.
Capability inventory: The skill uses Bash and Read tools. The associated CLI tool supports writing output to arbitrary files (--deliver file:<path>) and sending data to external webhooks (--deliver webhook:<url>).
Sanitization: The documentation does not specify any sanitization or validation of the retrieved API data before processing.
[DATA_EXFILTRATION]: The sutra-fitness-pp-cli tool includes a --deliver webhook:<url> flag. This capability allows the agent to POST command results—which include sensitive studio data such as client lists, contact information, and revenue reports—to any external URL.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of external binaries from the vendor's repositories. It uses npx to fetch the @mvanhorn/printing-press-library package and go install to fetch the sutra-fitness-pp-cli source from github.com/mvanhorn/printing-press-library. These are identified as legitimate vendor-owned resources.
[REMOTE_CODE_EXECUTION]: The installation procedures involve executing remote code via npx and compiling/installing remote source code via go install.

pp-sutra-fitness