validateAndResolveUrl resolved a hostname twice — once to validate, once to pin — leaving a TOCTOU window where DNS could answer differently between the two calls. The fix collapses both resolutions into one and pins the fetch to the entire validated address set instead of a single OS-picked IP.

How to write, store, and register a NeuroLink skill — JSON, frontmatter markdown, or a Claude-style SKILL.md directory — and how search_skills, list_skills, and the CLI's skills subcommand pick it up.

How dependents-scan.mjs scrapes GitHub's dependents graph with cheerio to sort real external users of @juspay/neurolink from Juspay's own repos and forks — and the unterminated shell quote that silently killed every gh CLI enrichment call it shipped with.

NeuroLink's first AI share-of-voice run recorded 0 of 10 answers citing it. The data shows less than that: four answers stop after one bullet, an 8B model read MCP and RAG as the Minecraft Protocol and a UI term, and the web probe extracted zero result URLs from the four pages it could parse.

On Gemini-3, a pure tool-calling turn emits thoughtSignature and functionCall parts with no text at all, so NeuroLink's prose-preservation branch never fired and the loop ignored the caller's abortSignal for 18 minutes and 74 tool calls.

How NeuroLink's TaskManager survives process restarts: BullMQ vs Node timers for scheduling, Redis vs filesystem for persistence, and a nine-phase autoresearch loop.

How NeuroLink's L2 embedding fast-path uses a hybrid cosine+BM25 retriever to skip the LLM router call for large tool catalogs, and how it had to be wired into the same code path that already applies session stickiness without breaking either mechanism.

How BaseServerAdapter, getFrameworkInstance(), and dynamic-import route groups let you mount NeuroLink MCP, proxy, and agent APIs onto a running Express or Fastify app without ceding server lifecycle control.

How NeuroLink models voice as TTS, STT, and full-duplex — ChunkedAudioStream backpressure, the realtime state machine, and OpenAI vs. Gemini Live divergence.

How NeuroLink's prepareRAGTool wires RAG into generate(): extension-based chunking strategy routing, FNV-1a hashing, a 128-dim InMemoryVectorStore, and a dynamically injected search_knowledge_base tool.