The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The SKILL.md file contains hardcoded credentials in a docker-compose.test.yml example, specifically using 'password' as the value for DATABASE_URL and POSTGRES_PASSWORD.
[PERSISTENCE]: The skill includes bash examples using cron syntax (0 20 * * * and 0 7 * * 1-5) to schedule the stopping and starting of AWS EC2 instances for cost optimization.
[COMMAND_EXECUTION]: The skill instructions demonstrate the use of CLI tools including docker-compose, npm, and the aws CLI to manage environment lifecycles and run tests.
[INDIRECT_PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection as it processes external environment configurations and infrastructure reports to perform management and parity checks.
Ingestion points: Ingests environment configurations, parity check results, and cost reports via the aqe/environment-management/ memory namespace as defined in SKILL.md.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands were identified in the ingestion workflow.
Capability inventory: The skill possesses the capability to execute shell commands via docker-compose exec, perform cloud operations using aws ec2, and execute application tests via npm test.
Sanitization: There is no evidence of sanitization or schema validation for the ingested environment data before it is processed by the agent.

test-environment-management