The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through the ingestion of trace data.
Ingestion points: Trace data is retrieved from OpenSearch indices such as otel-v1-apm-span-* and otel-v2-apm-service-map-* using PPL queries in traces.md and SKILL.md.
Boundary markers: The instructions lack delimiters or specific directives for the agent to distinguish between the trace data and its own system instructions.
Capability inventory: The skill has the ability to execute network requests via curl and specialized MCP tools like GenericOpenSearchApiTool.
Sanitization: There is no logic provided to sanitize or filter the content of the traces (e.g., span attributes or exception messages) before they are presented to the agent.
[SAFE]: The skill uses the vendor's own MCP server (opensearch-mcp-server-py) and manages credentials through environment variables, which is a recommended practice for secret management.
[SAFE]: The skill suggests disabling SSL verification in some configurations (using -k with curl or OPENSEARCH_SSL_VERIFY: "false"). This is a security best practice violation that could expose the connection to interception, but it is documented as a configuration option for self-managed or local environments.

trace-analytics