figma-to-jira
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill instructs the agent to generate and execute Python one-liner scripts at runtime to decode base64 image data and manage temporary files in the /tmp directory.
- Evidence: The use of python3 -c commands for base64 decoding and file operations defined in the Phase 2 workflow of SKILL.md.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted design metadata and context from Figma to populate Jira tickets, creating a surface where instructions embedded in the design could influence the agent.
- Ingestion points: get_metadata and get_design_context tools used in SKILL.md to fetch external design content.
- Boundary markers: Absent; there are no specific instructions to ignore embedded design data instructions.
- Capability inventory: The agent uses create_issue and update_issue for Jira and performs shell execution for file management.
- Sanitization: Absent; the skill does not define methods for filtering or escaping content retrieved from Figma.
- [COMMAND_EXECUTION]: The skill requires the use of shell commands for temporary file management and pipe operations to handle data transfers.
- Evidence: Use of shell pipes and python3 calls for file system tasks described in the Attaching Screenshots section of SKILL.md.
Audit Metadata