The Agent Skills Directory

[DATA_EXFILTRATION]: The skill provides a powerful output routing mechanism via the --deliver webhook:<url> flag. This allows the agent to automatically POST the results of any command—including those that retrieve user information, subscription details, or workspace secrets—to an arbitrary external URL.
[COMMAND_EXECUTION]: The skill relies on executing the elevenlabs-pp-cli binary via shell commands. It includes instructions for the agent to construct and execute these commands, which can manage various aspects of the ElevenLabs platform, including secret management and knowledge base updates.
[INDIRECT_PROMPT_INJECTION]: The skill exposes several surfaces where untrusted data from the web is ingested into the agent's context or the ElevenLabs platform.
Ingestion points: Commands such as convai create-url-document-route, convai add-documentation-to-knowledge-base, and audio-native update-content-from-url fetch and process content from external, potentially attacker-controlled URLs.
Boundary markers: The instructions do not define delimiters or warnings to treat ingested content as untrusted.
Capability inventory: The skill can execute shell commands, write to local files, and send data to webhooks.
Sanitization: No sanitization or validation of the fetched content is specified before it is processed or stored.
[EXTERNAL_DOWNLOADS]: The skill instructions facilitate the installation of external tools using npx and go install from the vendor's own repositories. These downloads are consistent with the skill's stated purpose and originate from the author's infrastructure.

pp-elevenlabs