The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches binaries and installer scripts from well-known services to provision the integration environment. This includes Elasticsearch from Elastic's official artifact repository, MinIO from its distribution site, and Node.js setup scripts from NodeSource and the official Node.js website. These downloads are from reputable sources and are used to provide standard development infrastructure.
[COMMAND_EXECUTION]: The skill uses various package managers and shell utilities to manage dependencies and system services. This involves executing apt-get for system-level tools, pip and uv for Python environment setup, and npm for Node.js package management. These commands are necessary for the skill's primary function of preparing a containerized agent environment.
[DATA_EXFILTRATION]: The skill is designed to manage environment variables and synchronize repository code. It includes workflows for configuring secrets via the veris-cli and pushing repository content to a remote environment for simulation. These operations are core functional requirements for integrating a repository with the Veris platform.
[PROMPT_INJECTION]: The skill processes user-provided repository files, including source code and configuration manifests, to generate integration logic. This represents an indirect prompt injection surface where data from the repository could contain instructions intended to influence the agent's behavior. Ingestion points: Repository source code, dependency manifests, and infrastructure configuration files (Phase 1). Boundary markers: Not identified; repository content is analyzed directly to inform configuration. Capability inventory: File system writing, environment variable management, and remote deployment via veris-cli. Sanitization: Not identified.

agent-integration