Stripe stripe-best-practices — agentic threat model

Not certain from the listing — relies on the host coding agent's underlying foundation model. Threats include prompt injection that could bypass security guidance or force the model to recommend deprecated, insecure, or malicious Stripe integration patterns.

The agent ingests authoritative Stripe documentation synced from docs.stripe.com/.well-known/skills. The primary threat is knowledge-base poisoning or man-in-the-middle attacks during the sync process, leading to compromised integration patterns being served to developers.

Acts as an 'Agent Skill' injected into a host framework. The framework must securely parse and isolate the SKILL.md instructions to prevent malicious instructions from hijacking the host agent's tool-calling or planning capabilities.

Not certain from the listing — deployment depends entirely on the host environment running the coding agent. If the host lacks sandboxing, executing code generated based on this agent's advice could expose local developer environments.

Not certain from the listing — there is no mention of built-in evaluation, logging, or guardrails to verify if the generated Stripe code actually adheres to the security practices (like restricted keys and webhook signatures) it recommends.

Focuses heavily on security best practices (restricted keys, webhooks, OAuth) as content, but the agent itself does not enforce compliance, authentication, or authorization controls on the code it generates or the developer using it.

Designed to be injected into other coding agents. A compromised Stripe skill could propagate vulnerabilities horizontally to multiple developer agents, leading to cascading security flaws across numerous downstream software integrations.