skills/afadtc/afa-dtc-skills/afa-paid/Gen Agent Trust Hub

afa-paid

Pass

Audited by Gen Agent Trust Hub on May 8, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE]: The skill documents the intention to write asset and campaign files to the ./brand-brain/ directory. This aligns with the described purpose of managing brand-specific advertising data and is performed within the expected workspace boundaries.
  • [PROMPT_INJECTION]: The skill defines an interface to receive user_request data from a central controller (Hub). This creates a surface for Indirect Prompt Injection (Category 8).
  • Ingestion points: Receives raw user_request strings in the input context (SKILL.md).
  • Boundary markers: The skill refers to _system/preamble.md and other protocol files for initialization, which typically include safety and boundary instructions.
  • Capability inventory: Performs file writes to the ./brand-brain/ directory and generates structured YAML outputs for the Hub.
  • Sanitization: Relies on external system-level protocols (_system/iron-rules.md, _system/interaction-protocol.md) for safety enforcement.
  • [EXTERNAL_DOWNLOADS]: No external downloads or remote dependencies are referenced in the instructions.
  • [COMMAND_EXECUTION]: No shell commands or direct system execution patterns were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
May 8, 2026, 12:58 AM
Security Audit — agent-trust-hub — afa-paid