The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data by mining recent agent conversation transcripts to identify user patterns.
Ingestion points: Transcript files located in the workspace-scoped path as defined in SKILL.md.
Boundary markers: The skill instructs the agent to "Mine bounded history" but lacks explicit instructions to sanitize or ignore potential prompt injection attacks embedded within those transcripts.
Capability inventory: The skill uses skill-creator to write new instruction files and performs Git operations such as commits and pull requests in Step 6.
Sanitization: No explicit sanitization or validation of transcript content is mentioned before it is interpolated into new skill instructions.
[EXTERNAL_DOWNLOADS]: The MIRROR.md file identifies the source of the skill as the official GitHub repository for Cursor plugins (https://github.com/cursor/plugins.git).

cursor-automate-me