The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill documentation encourages installing a CLI tool via curl -fsSL https://cli.inference.sh | sh. This pattern executes a remote script directly in the local shell, which is an unauthenticated and high-risk execution vector.
[EXTERNAL_DOWNLOADS]: The skill uses npx skills add to fetch and integrate additional functionality from inference-sh/skills at runtime. This introduces a dependency on external code hosted outside the primary skill package.
[COMMAND_EXECUTION]: The skill utilizes the infsh tool with broad permissions (Bash(infsh *)) to execute various AI-driven tasks. This tool performs network operations and interacts with external APIs to generate content and post to social media.
[PROMPT_INJECTION]: The skill is designed to ingest and reformat untrusted long-form content (such as blog posts or transcripts). This creates a vulnerability to indirect prompt injection if the source data contains malicious instructions intended to hijack the agent's behavior during the repurposing process.
Ingestion points: External text and audio data processed by the infsh tool via the provided command templates.
Boundary markers: There are no explicit delimiters or system instructions defined to prevent the agent from obeying commands embedded within the source content.
Capability inventory: The skill can execute shell commands (infsh) and perform social media actions (x/post-create).
Sanitization: No validation or escaping of the input content is implemented in the skill's markdown logic.

content-repurposing