The Agent Skills Directory

[DYNAMIC_EXECUTION]: The forecast.py script uses importlib.util to dynamically load and execute a Python module from the relative path ../pipeline/deals.py. Loading and executing code from a computed path outside of the skill's own directory is a security risk as it creates a dependency on the integrity of external files.
[INDIRECT_PROMPT_INJECTION]: The skill reads deal data (such as titles and stages) and interpolates it into the output provided to the AI agent. This establishes a vulnerability surface where an attacker could embed malicious instructions within deal metadata to influence the agent's behavior during the forecasting process.
Ingestion points: forecast.py loads deal data through the dynamically loaded pipeline module and iterates through the deals collection in the source file (e.g., pipeline.example.json).
Boundary markers: The output lacks explicit delimiters or instructions telling the agent to ignore any commands embedded within the deal details.
Capability inventory: The skill instructions encourage the agent to take further actions based on the data, such as using the /researcher or /pipeline move tools.
Sanitization: No sanitization or escaping of deal titles or reason strings is performed before they are printed to the agent's context.

forecaster