The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute several shell commands for testing, including pytest, npx ng test, npx jest, and a custom script scripts/smoke_turns.py. These commands are part of the standard testing workflow described.
[DYNAMIC_EXECUTION]: The agent is directed to use javascript_tool to inject and execute a "frame-capture snippet" on the development site https://dev.boisestate.ai. This constitutes runtime code execution within a browser context.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests data from external sources.
Ingestion points: The agent reads testing documentation (docs/testing/smoke-regression.md), test output files (/tmp/smoke.json), and live web content from https://dev.boisestate.ai via the read_page tool.
Boundary markers: The instructions do not define specific delimiters or warnings to ignore instructions embedded within the data being read.
Capability inventory: The skill has access to shell execution, file system operations, and browser-side JavaScript execution.
Sanitization: There is no mention of sanitizing or validating the ingested content before it is processed by the agent.

smoke-regression