The Agent Skills Directory

PROMPT_INJECTION (LOW): The skill possesses a surface for indirect prompt injection because it ingests untrusted image data and has file-writing capabilities.\n
Ingestion points: The skill ingests user-provided images via PIL.Image.open as described in SKILL.md and implemented in core/validators.py.\n
Boundary markers: Absent. There are no delimiters or instructions provided to the agent to ignore potentially malicious text or commands visually embedded in uploaded images.\n
Capability inventory: The skill can write files to the local filesystem using imageio.imwrite within the GIFBuilder.save method in core/gif_builder.py.\n
Sanitization: Absent. The skill does not perform any content-based sanitization or OCR-based filtering on ingested images.\n- SAFE (SAFE): No malicious behaviors such as credential exfiltration, unauthorized network connections, or persistence mechanisms were found. The skill utilizes well-known and trusted Python libraries.\n- LOW (LOW): The skill is currently incomplete or broken. Multiple template files (e.g., templates/flip.py, templates/pulse.py, templates/spin.py) reference a missing internal module core.typography. Additionally, functions such as draw_emoji and draw_emoji_enhanced are imported from core.frame_composer in various templates but are not actually defined in the provided core/frame_composer.py file.

slack-gif-creator