The Agent Skills Directory

[SAFE]: The skill consists entirely of instructional documentation and configuration for an AI agent to perform a code migration task. It does not contain executable code, remote script downloads, or network exfiltration patterns.
[DATA_EXPOSURE]: While the skill instructs the agent to audit repository code for markdown renderers and security properties, this is a standard developer workflow for library migration and does not involve accessing or exfiltrating sensitive credentials or personal data.
[INDIRECT_PROMPT_INJECTION]: The skill performs an audit of existing repository code, which is an ingestion point for untrusted data. However, the workflow is focused on structural code changes and documentation mapping, presenting a minimal attack surface that is mitigated by standard agent guardrails. No specific vulnerabilities were identified.

markstream-migration