clarion-sec-research
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script (
research.py) to perform indexing and search operations. These operations are restricted to specific subcommands (index,search,status) and operate on local data directories (~/clarion/queue and ~/clarion/sec). - [DATA_EXFILTRATION]: No unauthorized data transmission was detected. The script mentions a requirement for a
ZO_API_KEYin its error documentation, but does not handle or exfiltrate secrets itself. Networking is abstracted into vendor-provided libraries for the purpose of communicating with indexing services. - [REMOTE_CODE_EXECUTION]: The script uses internal vendor libraries (
ai_buffett_zo) to manage SEC data processing. There are no patterns of downloading and executing arbitrary code from external or untrusted sources. - [PROMPT_INJECTION]: The skill instructions define clear boundaries for the agent's behavior, emphasizing conservative interpretation of data and strict adherence to the indexed corpus. No instructions to bypass safety filters or override system prompts were found.
Audit Metadata