fp-smoke-test

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFEPROMPT_INJECTIONREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill incorporates user-provided input through the $ARGUMENTS variable without sanitization, allowing for potential direct prompt injection.
  • [COMMAND_EXECUTION]: The agent is instructed to identify and run build scripts, start development servers, and execute CLI commands found within the target application's source code.
  • [REMOTE_CODE_EXECUTION]: For library testing, the skill directs the agent to generate and execute arbitrary code snippets in the /tmp directory.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it relies on instructions parsed from local files like E2E_TESTS.md and README to determine its testing workflow.
  • Ingestion points: Files E2E_TESTS.md, README, source code, and the $ARGUMENTS variable.
  • Boundary markers: None; the skill lacks delimiters or warnings to treat ingested data as untrusted.
  • Capability inventory: The agent can execute shell commands, run generated scripts, perform network operations via curl, and use browser automation tools.
  • Sanitization: No filtering or validation is performed on the content extracted from the repository files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 09:22 AM