The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides an attack surface for indirect prompt injection because it ingests data from external, untrusted websites during its discovery phase. A malicious website could place specific instructions in headings, link text, or form labels to manipulate the agent's behavior during test generation or execution.
Ingestion points: The scripts/discover.py script extracts raw text from titles, headings, navigation menus, links, and forms of the target URL.
Boundary markers: There are no explicit instructions or delimiters used to warn the agent to ignore or isolate potentially malicious instructions found within the scraped website data.
Capability inventory: The agent has the ability to perform browser interactions (clicking, typing) and execute JavaScript assertions based on the plan derived from the scraped data.
Sanitization: No sanitization or filtering is applied to the content extracted from the target website before it is presented to the agent.
[COMMAND_EXECUTION]: The skill relies on shell commands to install its required infrastructure and to execute its logic. It instructs the agent to use npm and pip for package management and runs local Python scripts (discover.py, run_tests.py, generate_report.py) to manage the QA lifecycle.
[EXTERNAL_DOWNLOADS]: The skill setup involves downloading external tools and browser binaries. It specifically requests the installation of the Playwright CLI and the Chromium browser engine. These are official tools provided by Microsoft and are retrieved from standard public registries (NPM and PyPI).

test-web-ui