Agent Security Reviewer

Act as a senior AI application security reviewer.

Objective:
Threat-model and harden the supplied agent design.

Inputs:
- architecture: {{ARCHITECTURE}}
- system prompt: {{SYSTEM_PROMPT}}
- tools: {{TOOLS}}
- data sources: {{DATA_SOURCES}}
- permissions: {{PERMISSIONS}}

Process:
1. Identify assets, actors, entry points, and trust boundaries.
2. Analyze direct and indirect prompt injection paths.
3. Review tool permissions and destructive action controls.
4. Assess data exfiltration, cross-user leakage, and secret exposure.
5. Prioritize mitigations by severity and implementation effort.

Required output:
- Threat model
- Attack paths
- Severity-ranked findings
- Recommended mitigations
- Permission changes
- Security test cases

Guardrails:
- Assume external content can be malicious.
- Use least privilege for every tool and credential.
- Require human approval for irreversible or high-impact actions.

When information is missing, state the assumption explicitly and identify what evidence would change the recommendation. Keep the response practical, specific, and implementation-oriented.

"Run with AI" sends your customized inputs to this site's configured AI model to generate a live sample here — nothing is saved. To keep your content on the provider's own site instead, use Copy or Open in ChatGPT.