The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes internal Node.js scripts (install-mode.mjs, configure-telegram.mjs) to automate the installation of JSON configuration files and the setup of Telegram notification services. It also invokes the claude-mem CLI to manage worker service state.
[EXTERNAL_DOWNLOADS]: The skill initiates network communication with the official Telegram Bot API (api.telegram.org) to authenticate bot tokens and send test notifications. This is a well-known service used for the skill's primary alerting functionality.
[DATA_EXFILTRATION]: Observation metadata, including titles, projects, and IDs, are sent to the Telegram API when users enable alerts. The skill includes documentation alerting the user to the privacy implications of this data transfer.
[PROMPT_INJECTION]:
Ingestion points: The skill collects domain-specific taxonomy and behavior rules through interactive user interviews defined in SKILL.md.
Boundary markers: There are no explicit delimiters used to isolate user-provided text when it is interpolated into the generated mode's system prompts.
Capability inventory: The skill possesses capabilities for writing to the local filesystem (install-mode.mjs), modifying system configuration files, and performing network operations via the Telegram API.
Sanitization: While scripts validate structural integrity and ID formats, the descriptive text content for the generated prompts is not specifically sanitized against instruction injection.

mode-creator