The Agent Skills Directory

[COMMAND_EXECUTION]: The skill functionality is implemented through a local script scripts/ielts_buddy_api.py which is executed by the agent to perform API calls. This is a standard pattern for CLI-interfacing skills.
[INDIRECT_PROMPT_INJECTION]: The skill processes vocabulary data retrieved from an external API, creating a surface for potential injection if the data contains malicious instructions.
Ingestion points: Data is ingested through API search and card preparation functions within scripts/ielts_buddy_api.py (referenced in SKILL.md and references/vocabulary.md).
Boundary markers: The skill includes explicit instructions to obtain user confirmation before performing any write, delete, or import operations.
Capability inventory: The script provides capabilities for searching, adding, updating, deleting, and recording reviews of vocabulary data via shell execution.
Sanitization: The skill documentation specifies that only vocabulary content should be migrated, but no specific prompt delimiters or input sanitization logic is visible in the provided instruction files.
[EXTERNAL_DOWNLOADS]: The skill communicates with the vendor's API endpoint at https://work.ieltsbuddy.igopx.cn/api/v1/agent. This network activity is necessary for the skill's stated purpose of vocabulary management and synchronization.

ielts-vocabulary-coach