Claude Skills by maruakshay

Review an AI system for credentials leaking into prompt context — covering API keys in system prompts, database connection strings in RAG chunks, access tokens in tool arguments, and exfiltration of secrets from context via prompt injection.

Review a Semantic Kernel agent or planner for unauthorized plugin execution, unsafe function routing, excessive connector permissions, missing argument validation, insecure memory use, and weak sandboxing around code or action execution.

Review synthetic training data generation pipelines for distillation attack risks, model collapse vectors, privacy leakage through synthetic outputs, and quality controls that prevent degraded or adversarially biased synthetic data from entering training pipelines.

Review the system and infrastructure around an AI application for secret handling, network exposure, runtime isolation, model and dependency supply chain risk, weak observability controls, and deployment misconfiguration across cloud and self-hosted environments.

Assess the security and operational risk of third-party model APIs and vendors — covering due diligence requirements, API dependency assessment, data processing agreements, vendor lock-in risk, SLA gaps for AI-specific failure modes, and contingency planning for vendor model changes.

Review tool definition schemas in AI agent systems for malformed definitions, schema injection, parameter confusion attacks, and tool name collisions that cause agents to invoke unintended tools or with manipulated arguments.

Review an agentic AI system for insecure tool calling, unsafe code or shell execution, excessive permissions, confirmation bypass, argument injection, sandbox escape risk, and action-policy failures across languages and frameworks.

Review training datasets for backdoor triggers, label flipping attacks, targeted behavioral modification, and data collection pipeline vulnerabilities that allow adversarial influence over model behavior at inference time.