The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute ctx7 CLI commands for managing skill configurations and installations.
[EXTERNAL_DOWNLOADS]: The skill fetches content from external repositories via the ctx7 skills install and ctx7 skills info commands as part of its core functionality.
[REMOTE_CODE_EXECUTION]: The skill facilitates the installation of external 'skills' from remote repositories. Since these skills contain instructions and potential scripts for the agent, this process constitutes a form of remote code execution intended by the skill's design.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to the way it handles external data.
Ingestion points: Untrusted metadata (such as skill names and descriptions) is ingested from external repositories during search and info operations.
Boundary markers: There are no boundary markers or instructions to ignore embedded commands; the skill instructions specifically require preserving entry text 'as-is'.
Capability inventory: The skill environment includes Bash execution and WebFetch capabilities.
Sanitization: No sanitization or validation is applied to data retrieved from remote sources before it is displayed to the user or processed by the agent.

context7-skills