The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user and agent to install the belt-sh/cli package via npx. It also references installation instructions hosted on a remote GitHub repository (https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md).
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute belt commands, such as belt login and belt app run. These commands involve network operations and authentication. If the agent interpolates user-provided text directly into these shell commands without proper sanitization, it could lead to command injection.
[INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it ingests data from external search tools (tavily/search-assistant, exa/search).
Ingestion points: Data enters the context via the output of belt app run commands (in SKILL.md).
Boundary markers: None present; external content is processed directly to fact-check claims.
Capability inventory: The skill possesses shell execution capabilities (Bash tool) and network access through the belt CLI.
Sanitization: No explicit sanitization or filtering of the external search results is defined in the instructions.

press-release-writing