configuring-topics
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents the setup of Topic Listeners (TopicExport) which ingest asynchronous message payloads from external sources. This defines an architectural ingestion surface where external content is incorporated into the platform's processing environment.
- Ingestion points: External JSON messages are received via TopicExport steps as described in SKILL.md and references/cli.md.
- Boundary markers: The configuration steps do not specify the use of delimiters or safety instructions (like 'ignore embedded instructions') for the content of the message payloads.
- Capability inventory: The skill provides tools for writing and updating flows (upsert_flow), exports (upsert_export), and imports (upsert_import) that process this data.
- Sanitization: Sanitization procedures for message content are not described in these infrastructure configuration instructions.
- [COMMAND_EXECUTION]: The skill provides a comprehensive set of celigo CLI commands in references/cli.md for managing topics, messages, and integrations. These are standard operational commands for the platform's command-line interface.
Audit Metadata