The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, which could contain malicious instructions.
Ingestion points: External content is ingested via infsh/agent-browser (e.g., in the 'Media in Threads' section) and tavily/search-assistant (e.g., in the 'Repurposing to Thread' section).
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the fetched web content or search results.
Capability inventory: The skill has the capability to post content to X (x/post-create), generate images (infsh/html-to-image), and execute shell commands via the belt CLI.
Sanitization: No explicit sanitization or filtering of the external content is mentioned before it is processed by the agent to create thread content.
[EXTERNAL_DOWNLOADS]: The skill references and directs the agent to fetch installation instructions from the vendor's GitHub repository (https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md).
[COMMAND_EXECUTION]: The skill uses the belt CLI tool via Bash to perform actions such as logging in, posting tweets, browsing the web, and performing searches. This is part of the skill's defined configuration in allowed-tools.

twitter-thread-creation