Connectivity, chat, JSON & streaming regression tests for all enabled cloud LLM providers
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
At runtime, the required workflow makes network calls to third-party model/provider endpoints and ingests their returned message text (including any free-form content in `resp.text()`/SSE `data:` chunks) via `readSSEStream(...)` / `resp.json()` into variables like `content`, which are then included in emitted JSONL events (and thus into the agent runtime context).
SharpAI/DeepCamera
933dcc7
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.