The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is built around executing the te-cli command-line utility. It maps natural language requests to specific CLI commands such as te-cli analysis +query-sql and te-cli analysis +save-report.\n- [INDIRECT_PROMPT_INJECTION]: The skill exposes a significant attack surface by allowing the execution of arbitrary SQL queries and the processing of complex JSON configurations. If the agent uses data from untrusted external sources to construct these parameters, it could be manipulated into performing unauthorized data access or resource modification.\n
Ingestion points: Untrusted data can enter via the --sql parameter in references/query-sql.md, and via --events, --event-view, or --reports JSON parameters in references/save-report.md and references/update-dashboard.md.\n
Boundary markers: The skill instructions do not specify any delimiters or boundary markers to distinguish user-provided data from system instructions when interpolating parameters into the CLI commands.\n
Capability inventory: The skill possesses the capability to read sensitive data via SQL and WebSocket queries, and to modify platform state by creating or updating reports and dashboards.\n
Sanitization: There is no evidence of input validation, escaping, or sanitization within the skill's instructions to prevent command or SQL injection attacks.

te-analysis