The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill instructs the agent to maintain a run history by executing a Node.js one-liner via the shell command node -e. This script performs local file system reads and writes to a JSON file (run-history.json) located within the skill's internal directory structure. This constitutes runtime script generation and execution for local state management.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and assess external source code, creating a vulnerability surface where untrusted input could attempt to influence the agent's summary, scores, or subsequent recommendations.
Ingestion points: External source files in directories specified by the --scope argument (e.g., src/).
Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded prompts within the analyzed code.
Capability inventory: The skill utilizes subprocess execution (via the aqe CLI) and file system modification capabilities.
Sanitization: There is no evidence of sanitization or strict schema validation for the raw code content before it is processed by the analysis agents.
[COMMAND_EXECUTION]: The skill relies on the execution of a specialized CLI tool (aqe) to perform its primary functions, including quality assessment, deployment readiness checks, and report generation.

qe-quality-assessment