The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to perform server discovery, manage Docker containers, and generate cryptographic secrets using openssl. It also uses the mmctl CLI tool to modify Mattermost server configurations.
[DYNAMIC_EXECUTION]: Node.js is invoked using the -e flag to perform runtime environment checks, such as verifying if port 8065 is available and inspecting package.json for specific dependencies.
[INDIRECT_PROMPT_INJECTION]: The skill configures a Mattermost adapter that ingests data from external chat channels, which is a potential vector for indirect prompt injection.
Ingestion points: Untrusted data enters the agent context through the Mattermost WebSocket adapter located at src/channels/mattermost.ts.
Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within incoming messages.
Capability inventory: The agent environment has access to privileged tools including docker, pnpm, and file system write operations as seen in SKILL.md and scripts/discover-server.mjs.
Sanitization: There is no evidence of message content sanitization or escaping within the setup scripts.

add-mattermost