skills/capire/skills/cap-developer/Gen Agent Trust Hub

cap-developer

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: No security issues detected. The skill provides development guidelines for SAP CAP, adhering to best practices and discouraging unsafe behaviors such as hardcoding credentials.
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read existing project files (CDS models) to provide context-aware advice. While this represents a theoretical attack surface for indirect prompt injection, it is standard for coding assistant skills and the skill does not grant the agent high-privilege autonomous capabilities that could be easily exploited.
  • Ingestion points: Local project files (CDS models) read via the CAP MCP server as specified in SKILL.md.
  • Boundary markers: Absent in the provided instructions.
  • Capability inventory: Generation and execution of 'cds' and 'mvn' commands, as well as file system modifications for project setup and custom handler development (SKILL.md, references/java.md, references/nodejs.md).
  • Sanitization: No explicit sanitization or validation of the input CDS models is described before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 08:14 PM
Security Audit — agent-trust-hub — cap-developer