indykite-authzen-search-action

Pass

Audited by Gen Agent Trust Hub on Sep 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill includes a shell script scripts/search-action.sh that is intended to be executed to perform API requests. The script contains safety measures such as IndyKite-specific host pinning and JSON format validation.
  • [DATA_EXFILTRATION]: The skill performs network requests to eu.api.indykite.com and us.api.indykite.com. These operations are consistent with the skill's primary purpose of communicating with the vendor's API.
  • [INDIRECT_PROMPT_INJECTION]: The scripts/search-action.sh script ingests data from a user-supplied file. * Ingestion points: scripts/search-action.sh reads a file path provided as a command-line argument. * Boundary markers: The skill documentation lacks explicit instructions for the agent to treat this data as untrusted or to use specific delimiters when processing it. * Capability inventory: The script uses curl to transmit data to a remote endpoint. * Sanitization: The script performs a structural check using jq to ensure the input is valid JSON, but does not sanitize the content within the JSON fields.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 24, 2026, 07:09 AM
Security Audit — agent-trust-hub — indykite-authzen-search-action