The Agent Skills Directory

[DATA_EXFILTRATION]: The skill accesses the sensitive local configuration file ~/.opik.config and the environment variable OPIK_API_KEY to authenticate its requests to the Opik platform. This is consistent with the skill's primary purpose of diagnosing traces.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes trace data from external Opik SDK calls, creating a surface for indirect prompt injection.
Ingestion points: Data is retrieved via client.search_traces() and client.rest_client.agent_insights.find_agent_insights_issues() within SKILL.md.
Boundary markers: The instructions do not define delimiters or specific warnings to ignore instructions embedded within the trace data.
Capability inventory: The skill utilizes the Python SDK for data access and specifies Bash in its allowed-tools for execution.
Sanitization: No explicit sanitization or filtering of the retrieved trace content is described in the diagnostic logic.

opik-diagnose