The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, such as source code files and project plans, which serves as a potential vector for indirect prompt injection.
Ingestion points: The scripts/complexity_scan.py script reads source code files from the repository root; reduce.md reads user-provided plans or specification files.
Boundary markers: There are no explicit delimiters or instructions for the agent to ignore embedded commands within the ingested data.
Capability inventory: The skill can execute local Python scripts (complexity_scan.py), write or update documentation files (CONTEXT.md, DEFINED_TERMS.md, and ADRs), and invoke other agent capabilities or skills such as codex-exec, claude, and visualize.
Sanitization: No specific sanitization or escaping of the ingested code or plan content is performed before it is interpolated into the agent's context.
[COMMAND_EXECUTION]: The skill executes a bundled Python script (scripts/complexity_scan.py) to generate code analysis reports. This script uses standard Python libraries to perform static analysis through AST parsing and regular expressions without performing network operations.

engineering