budget-optimizer

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted user data (e.g., campaign goals, audience descriptions) to generate budget plans.
  • Ingestion points: User input fields defined in Step 1 (Campaign Goal, Target Audience, Timeline, Geographic Focus).
  • Boundary markers: None specified in the instructions to separate user data from instructions.
  • Capability inventory: File system write access to the memory/ directory via markdown output.
  • Sanitization: No explicit sanitization or validation of user-provided strings before interpolation into the budget templates.
  • [DATA_EXFILTRATION]: The skill includes instructions to save budget recommendations and campaign facts to a local memory/ directory (e.g., memory/influencer/budget-optimizer/YYYY-MM-DD-<topic>.md and memory/hot-cache.md). This is standard session persistence for the platform and does not involve sending data to external, non-whitelisted domains.
  • [SAFE]: No executable code, remote script downloads, or obfuscation techniques were detected in the files. All external links point to the author's official GitHub repository or shared platform documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 09:34 AM
Security Audit — agent-trust-hub — budget-optimizer