rhdh-jira-sprint-plan
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from Jira work items (summaries, descriptions, etc.) which are external inputs. This is a standard attack surface where malicious content in Jira could attempt to influence the agent's behavior.
- Ingestion points: Jira search results retrieved via
acli jira workitem searchinworkflows/plan-sprint.md(Steps 2, 3, 5, 8, 9). - Boundary markers: The workflow lacks explicit boundary markers or instructions to treat Jira content as untrusted data.
- Capability inventory: The skill executes
aclicommands and interacts with other internal workflows (/rhdh-jira-api,/rhdh-jira-update). - Sanitization: There is no evidence of sanitization for the retrieved Jira data.
- [COMMAND_EXECUTION]: The workflow relies on the
aclicommand-line tool to perform read operations on Jira boards and work items. These operations are restricted to listing and searching, matching the read-only intent described in the skill documentation.
Audit Metadata