orbitant-graceful-degradation
Pass
Audited by Gen Agent Trust Hub on Aug 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to ingest and process data from various external and untrusted sources, which exposes the agent to potential indirect prompt injection attacks.
- Ingestion points: The skill retrieves data from multiple MCP tools (Gmail, Google Calendar, Slack, Asana) and reads local filesystem content from project-specific and global configuration files (
rocks.yaml,voice.md,constraints.yaml, and thecontacts/directory). - Boundary markers: There are no explicit instructions or delimiters defined to isolate the content retrieved from these sources or to instruct the agent to ignore any commands potentially embedded within the emails, slack messages, or configuration files.
- Capability inventory: The skill performs read-only operations and connectivity probes. It does not contain instructions for arbitrary code execution, subprocess spawning, or direct network writes based on the retrieved data, which limits the potential impact of an injection.
- Sanitization: The instructions do not specify any validation, filtering, or escaping of the content fetched from the MCP tools or filesystem before it is used to generate the availability report.
Audit Metadata