ib-pmcc-advisor
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script (
pmcc_advisor.py) usinguv runto interface with Interactive Brokers and perform financial calculations. This execution is scoped to the analysis of PMCC positions and uses standard CLI arguments. - [INDIRECT_PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection because it ingests and processes external data to generate advice.
- Ingestion points: Interactive Brokers portfolio data and Yahoo Finance earnings information are retrieved via the analysis script.
- Boundary markers: The instructions do not specify explicit delimiters to isolate the ingested financial data from the prompt instructions.
- Capability inventory: The skill has the ability to execute local Python scripts and write analysis reports or JSON data to the
sandbox/directory. - Sanitization: External data is fetched and structured as JSON by the Python script before being presented to the agent, providing a layer of validation against unstructured injection.
Audit Metadata