launch-sync-environment-variables-from-env-example

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill parses data from a local .env.example file to identify environment variable keys for synchronization, which presents an ingestion surface for untrusted local content. \n
  • Ingestion points: The skill reads and parses the content of the user-provided .env.example file (SKILL.md). \n
  • Boundary markers: There are no explicit delimiters or boundary markers specified for isolating the file content during the parsing process. \n
  • Capability inventory: The skill uses the Bash tool, the Read tool, and performs PATCH requests to the Contentstack Launch API (SKILL.md). \n
  • Sanitization: Instructions mandate ignoring comments, blank lines, and malformed entries, relying on the agent's logic to safely extract keys without executing embedded instructions. \n- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute synchronization logic, perform file operations, and interact with the Contentstack Launch API. \n- [DATA_EXFILTRATION]: The skill transmits environment variable keys to the Contentstack Launch API (projects/{project_uid}/environments/{environment_uid}). \n
  • Evidence: This interaction targets the official Contentstack platform to synchronize configuration. The instructions explicitly forbid the transmission, logging, or exposure of secret values or tokens. \n- [EXTERNAL_DOWNLOADS]: The skill interacts with the external Contentstack Launch API to fetch current environment states and update configurations. \n
  • Evidence: Network operations are directed at the official API endpoints for the Contentstack platform to facilitate environment management.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 09:43 PM
Security Audit — agent-trust-hub — launch-sync-environment-variables-from-env-example