The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data, including user-supplied Blender files, images, and results from external image searches.
Ingestion points: SKILL.md (processing existing Blender candidates) and references/astra-workflow.md (retrieving and inspecting images via search tools).
Boundary markers: The instructions do not specify the use of clear delimiters or instructions to ignore embedded commands within external data (e.g., in image metadata or search results).
Capability inventory: The skill has access to the Blender executable, file-writing capabilities (iteration_review.json, work_state.md), and network access for asset searches and downloads.
Sanitization: There is no mention of sanitizing or validating external content before processing.
[DYNAMIC_EXECUTION]: The skill requires the agent to generate and execute Python code to perform targeted repairs on Blender source files.
Evidence: SKILL.md directs the agent to "Make one targeted repair... in the durable Python or editable Blender source" and then "Regenerate the authored and exported artifacts." This represents a dynamic code generation and execution surface necessary for the skill's primary purpose.

blender-iterative-refinement