pp-tella

Warn

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches the CLI tool from the vendor's GitHub repository and via the NPM registry during the prerequisite installation phase.
  • [COMMAND_EXECUTION]: The skill uses shell commands to install the CLI and perform video, playlist, and webhook operations.
  • [CREDENTIALS_UNSAFE]: The documentation instructs users to manually capture and export raw browser session cookies (__Secure-Tella.session) to environment variables to enable 'unofficial' AI-driven features. This practice exposes sensitive session tokens to the environment where they can be accessed by the tool and the agent.
  • [DATA_EXFILTRATION]: The tool includes a --deliver webhook:<url> flag that allows the agent to POST command results, which may contain sensitive data like video transcripts or workspace metadata, to any user-specified external URL.
  • [PROMPT_INJECTION]: The skill processes untrusted data by searching through video transcripts (Ingestion point: transcripts search command in SKILL.md). There are no explicit boundary markers or sanitization mentioned. Combined with the agent's capabilities to write files and perform network operations via the --deliver flag, this creates a surface for indirect prompt injection attacks if malicious content is embedded in transcripts.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 22, 2026, 03:01 AM
Security Audit — agent-trust-hub — pp-tella