The Agent Skills Directory

[COMMAND_EXECUTION]: The skill employs a custom shell script scripts/tfy-api.sh to facilitate authenticated REST API requests to the TrueFoundry platform. It also uses the tfy command-line interface for resource configuration and application.\n- [EXTERNAL_DOWNLOADS]: Documentation within the skill provides instructions for installing the official truefoundry Python package (version 0.5.0), which is the primary tool for interacting with the vendor's platform.\n- [PROMPT_INJECTION]: The skill identifies potential attack surfaces for Indirect Prompt Injection (Category 8) and provides specific mitigations for the AI agent.\n
Ingestion points: The gateway can register external sources such as A2A agent cards (agent_card_url) and OpenAPI specifications (spec.url).\n
Boundary markers: Instructions in SKILL.md and references/manifest-schema.md require the agent to obtain explicit human confirmation before connecting to or onboarding external URLs.\n
Capability inventory: The skill performs sensitive operations via tfy apply and direct REST API calls through the tfy-api.sh script.\n
Sanitization: Guidelines emphasize verifying endpoint authenticity and using TrueFoundry's secret management for authentication headers.\n- [SAFE]: No malicious code patterns, obfuscation, or unauthorized data access were found. The skill correctly uses placeholders and secret references for sensitive configuration, ensuring that credentials are not hardcoded or exposed.

truefoundry-ai-gateway