jira-qa-scope
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources to generate its output.
- Ingestion points: The skill ingests Jira ticket summaries, descriptions, and comments, along with Git commit messages and code diffs (SKILL.md, Steps 2 and 3).
- Capability inventory: The skill has the capability to write Markdown files to the local docs/ directory and post comments to Jira issues through MCP or CLI tools (SKILL.md, Steps 4 and 6).
- Boundary markers: There are no explicit prompt boundary markers or sanitization instructions defined to isolate external data from the skill's internal logic.
- Sanitization: The risk of malicious content in tickets or commits influencing the agent is mitigated by a mandatory user review loop (Step 5) and the requirement for explicit user approval before any external Jira comments are posted (Step 6).
- [COMMAND_EXECUTION]: The skill executes local command-line tools including git (fetch, log, show, diff) and gh (pr view, pr diff, api) to analyze repository history and metadata. These are used as intended for the skill's core functionality of analyzing development changes for QA planning.
Audit Metadata