The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's workflow depends on reading and analyzing external data that could be controlled by an attacker to influence the agent's behavior.
Ingestion points: The skill reads external requirements documents (.specs/SPEC-*.md) and content from linked GitHub Issues in its 'Requirements Analysis' phase (SKILL.md).
Boundary markers: There are no explicit instructions, such as XML tags or 'ignore embedded instructions' markers, to prevent the agent from interpreting data within those external sources as new system-level commands.
Capability inventory: The skill possesses the ability to execute shell commands for testing purposes (dotnet test, Playwright) and can perform write actions to GitHub repositories via the /create-issues skill.
Sanitization: The skill lacks specified sanitization or validation logic for data retrieved from external sources before using that data to generate test plans or execute automated scripts.
[COMMAND_EXECUTION]: The skill requires the agent to execute automated test suites using development tools like dotnet test and Playwright. While these are standard QA activities, they involve executing code based on the contents of the project's codebase, which could be exploited if the codebase contains malicious test definitions.

qa-analyst