Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward wrapper for a configured stock-analysis REST API, with its network use and optional chat endpoint disclosed in the artifact.
Before installing, confirm that DSA_BASE_URL points to a DSA service you trust. If it is remote, stock codes, analysis prompts, optional chat messages, session identifiers, and any required cookies may be sent to that service. Avoid sending private financial details unless you control or trust the endpoint.
The skill description does not warn users that their stock query will be transmitted to an external REST API endpoint defined by DSA_BASE_URL. This creates a transparency and privacy problem because users may reasonably believe analysis is local while their query content is sent to another service, potentially including remote infrastructure.
The trigger condition is broad enough that normal user requests like 'analyze AAPL' or stock-related discussion may invoke the skill without clearly signaling that data will be sent to an external service. This can cause unintended tool use and external disclosure of user-provided financial queries, especially when the user did not explicitly consent to API-backed analysis.
The optional agent chat workflow explicitly sends free-form user messages and a session identifier to an external endpoint, which can expose sensitive financial interests, conversation context, or identifiers to the DSA service. The risk is increased because the skill notes that remote deployment is supported and authentication may be absent by default, so user content could be transmitted to infrastructure outside the local trust boundary.
若 DSA 已启用 AGENT_MODE=true，可调用 Agent 策略问股接口，支持多轮对话与多种策略（缠论、均线金叉等）：
curl -X POST {DSA_BASE_URL}/api/v1/agent/chat \
-H 'Content-Type: application/json' \
-d '{"message": "用缠论分析 600519", "session_id": "optional-session-id"}'
The description and examples are primarily written to handle Chinese requests and state support limitations in Chinese, while not documenting any user choice of language or locale behavior. This can be read as imposing a specific language/locale pattern without explicit opt-in.
No suspicious patterns detected.