jira-cr
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external data from Jira (issue summaries, descriptions, acceptance criteria, and comments) and Git repositories (diffs, patches, and source code). This content is used to inform the agent's summary generation, creating a surface where malicious instructions within a ticket or code could attempt to influence the agent's behavior.
- Ingestion points: Jira issue retrieval and comment fetching in Step 1; Git status, diff, and repository inspection in Step 3.
- Boundary markers: The skill utilizes a structured markdown template for its output (Step 5), but lacks explicit delimiters or instructions to treat external data as untrusted content.
- Capability inventory: Jira tools for site discovery and issue/comment management (read/write); Git CLI for repository inspection (read).
- Sanitization: The skill includes a specific instruction to 'Exclude unrelated changes and sensitive values' when reading code changes (Step 3).
- [COMMAND_EXECUTION]: The skill executes local Git commands to resolve the scope of changes.
- Evidence: Use of
git diff --cached,git status, and inspection of patches and branches as described in Step 3. - Context: These commands are limited to read-only inspection of the project context and are essential for the skill's primary functionality.
Audit Metadata