The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill's installation instructions include the command npx clawhub install portfolio-risk-manager. The use of npx to execute a package from a non-standard source like clawhub (which is not a recognized registry or trusted service) introduces the risk of downloading and executing unverified external code.
[COMMAND_EXECUTION]: The skill relies on the execution of local Python scripts (scripts/portfolio_risk.py and scripts/build_optimizer_inputs.py) to analyze portfolio data. While these scripts perform legitimate mathematical operations, they execute in the local environment and process user-supplied file paths.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes untrusted data from external JSON files.
Ingestion points: Financial data ingested from assets/my_holdings.json and assets/historical_returns.json as specified in the command examples.
Boundary markers: None; there are no delimiters or instructions to the agent to treat the content of the data files as untrusted or to ignore embedded instructions.
Capability inventory: The skill has the capability to read local files and execute Python scripts, which could be leveraged if an injected instruction influences the agent's next steps.
Sanitization: The Python scripts do not perform any sanitization or validation of the string fields (e.g., ticker, sector, region) before they are output and read by the agent.

portfolio-risk-manager