The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources which could potentially contain malicious instructions intended for the underlying LLM.
Ingestion points: The skill accepts untrusted input through the materialText, materialFile, and materialUrl parameters as defined in references/form-schema.json and scripts/run.py.
Boundary markers: No explicit boundary markers or delimiters are defined in the instructions to isolate user-provided content from the system prompt.
Capability inventory: The provided scripts/run.py runner is limited to performing network requests to the vendor API and printing the result. It does not execute local shell commands or write files based on the API response.
Sanitization: No input sanitization or validation logic is present in the provided client-side scripts.
[DATA_EXFILTRATION]: The skill transmits user-provided data to an external endpoint for processing.
Evidence: The script scripts/run.py sends the params object containing user data to https://ai-skills.ai/api/execute using a POST request.
Context: The destination domain ai-skills.ai is the official service domain for this skill as documented in SKILL.md and matches the vendor's identified infrastructure.

ai-core-web-vitals