The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze telemetry data from Elasticsearch that originates from external LLM interactions.
Ingestion points: Data is pulled from traces-*, metrics-*, and logs-* data streams via POST /_query commands as detailed in SKILL.md and references/esql-recipes.md.
Boundary markers: There are no explicit instructions or delimiters defined to isolate the agent from potentially malicious instructions embedded within the monitored LLM prompt or response logs.
Capability inventory: The skill possesses the capability to execute complex ES|QL queries and inspect cluster metadata using the elastic CLI.
Sanitization: The skill relies on structured ES|QL for data extraction but does not specify sanitization or filtering of the actual content retrieved from the telemetry streams.
[COMMAND_EXECUTION]: The skill relies on the elastic CLI to interact with Elasticsearch and Kibana. It maps various HTTP shorthands to shell commands such as elastic es info, elastic es indices get-mapping, and elastic es esql query.
[EXTERNAL_DOWNLOADS]: The skill refers to the official GitHub repository for the elastic CLI (github.com/elastic/cli) to guide users on environment configuration.

observability-llm-obs