jira-tickets
Fail
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: HIGHPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTION
Full Analysis
- [PRIVILEGE_ESCALATION]: The skill explicitly instructs the agent to re-run commands with escalated permissions if a security sandbox blocks execution (e.g., when receiving a 'runtime_access_denied' error). This encourages bypassing security boundaries.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from Jira tickets, comments, and labels, which could contain malicious instructions.
- Ingestion points: Jira ticket content (SKILL.md).
- Boundary markers: Includes instructions to treat ticket text as data, not instructions.
- Capability inventory: Reading and updating Jira tickets via the Orca CLI (SKILL.md).
- Sanitization: No explicit sanitization or validation mentioned.
- [COMMAND_EXECUTION]: The skill dynamically determines which executable to run based on environment variables such as 'ORCA_CLI_COMMAND' and 'ORCA_DEV_REPO_ROOT'.
- [EXTERNAL_DOWNLOADS]: The skill uses the 'orca skills get' command to fetch and load additional ticket workflows and command definitions at runtime.
- [DYNAMIC_EXECUTION]: The skill resolves the binary path at runtime and loads dynamic skill content to extend its functionality.
Recommendations
- AI detected serious security threats
Audit Metadata