skills/camunda/skills/camunda-feel/Gen Agent Trust Hub

camunda-feel

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides instructions for using the c8ctl command-line utility to evaluate and debug FEEL expressions. This is a standard development and operational task for the Camunda platform.
  • [DYNAMIC_EXECUTION]: FEEL expressions are dynamically evaluated using either a local JavaScript-based engine (feelin) or a remote Scala-based cluster engine. This dynamic evaluation is the intended primary purpose of the skill.
  • [INDIRECT_PROMPT_INJECTION]: The skill documents the fromAi() function, which defines parameters meant to be populated by an LLM, creating an interface for AI-generated data ingestion.
  • Ingestion points: FEEL expressions and variable contexts passed to the c8ctl feel evaluate command in SKILL.md.
  • Boundary markers: None explicitly defined in the instructions.
  • Capability inventory: Evaluation of arithmetic, logical, and context-based expressions, and execution of the c8ctl CLI in SKILL.md.
  • Sanitization: The skill describes type validation and warnings for functional correctness but does not include specific security-focused sanitization for the expressions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 06:12 PM
Security Audit — agent-trust-hub — camunda-feel