The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill contains a hardcoded password "My_password_123!@#" for the OpenSearch admin user in the "Connection Defaults" section of SKILL.md. Hardcoding credentials, even as defaults for local environments, poses a security risk as they may be inadvertently used in real environments.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests data from external APIs and interpolates it into the agent context.
Ingestion points: Metric data and query results are fetched from the Prometheus HTTP API (including "/api/v1/query", "/api/v1/alerts", and "/api/v1/rules") throughout SKILL.md.
Boundary markers: There are no specific boundary markers or instructions to the agent to ignore potentially malicious instructions within the Prometheus responses.
Capability inventory: The skill uses "Bash" and "curl" across multiple operations, which allows for network requests and local command execution.
Sanitization: The skill lacks sanitization or validation logic for data retrieved from external endpoints before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill references documentation and configuration rules from the opensearch-project GitHub repository. These are documented neutrally as references for implementing SLI/SLO queries.
[COMMAND_EXECUTION]: The skill uses "curl" within "Bash" commands to perform operations including querying the Prometheus API and triggering configuration reloads via the "/-/reload" endpoint.

slo-sli