The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill enables the agent to process data originating from untrusted production environments through the EAS Observe telemetry system.\n
Ingestion points: Telemetry data, including custom event attributes and error messages, is pulled into the agent context via eas observe:events and observe_events tools (see references/queries.md).\n
Boundary markers: No specific instructions or delimiters are provided to the agent to distinguish between telemetry data and instructions.\n
Capability inventory: The agent can execute CLI commands and MCP tools to retrieve and analyze logs.\n
Sanitization: The documentation focuses on PII sanitization for data upload but lacks guidance for the agent to sanitize retrieved strings during the analysis phase.\n- [EXTERNAL_DOWNLOADS]: The skill references the installation and execution of external packages.\n
Evidence: Commands such as npx expo install expo-observe and npx --yes @expo/agent-cli@latest are used for setup and feedback (documented in SKILL.md and references/setup.md).\n
Note: These resources are hosted on official registries under the author's verified namespace.\n- [COMMAND_EXECUTION]: The skill relies on the execution of various CLI tools for its core functionality.\n
Evidence: Extensive documentation of eas observe:* commands in references/queries.md.\n
Note: This behavior is aligned with the skill's primary purpose of performance monitoring and error reporting.

eas-observe