telnyx-storage-curl

Warn

Audited by Gen Agent Trust Hub on Mar 26, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides multiple bash command templates using curl to interact with the Telnyx API, including operations for certificate management and usage tracking.
  • [DATA_EXFILTRATION]: The skill provides instructions for the agent to read local sensitive files and transmit them over the network. Specifically, the 'Add SSL Certificate' section includes a curl command using the @ syntax (-F "private_key=@/path/to/file") to upload a private key to https://api.telnyx.com/v2/storage/buckets/{bucketName}/ssl_certificate. This pattern can be abused to exfiltrate unintended sensitive files from the host system.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it interpolates untrusted data into shell commands.
  • Ingestion points: Data is ingested into command placeholders such as {bucketName}, {objectName}, and /path/to/file within the SKILL.md file.
  • Boundary markers: No boundary markers or specific instructions are provided to the agent to treat interpolated content as untrusted or non-executable.
  • Capability inventory: The skill utilizes curl, which grants the agent both network access and the ability to read arbitrary files from the filesystem.
  • Sanitization: There is no evidence of input validation, escaping, or sanitization for the parameters before they are executed in a shell context.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 26, 2026, 05:52 PM