The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
Tavily cache and finance-mcp
Local MCP server used by the deep-research agent, plus the on-disk search cache.
API keys are not included. mcp/example.env has empty placeholders. Copy it to .env and fill in your own keys before starting the server.
mcp/
finance-mcp source. cache/, logs, .env, and key files are not in this folder.
tool_cache/ and extract_ths_tool_cache/ are the server's bundled tool caches.
cache/
One gzipped JSONL file per tool. Each line is one cache entry:
{"key": "<cache key>", "meta": {"created_time": "...", "data_type": "dict"}, "data": {}}
| file | entries |
|---|---|
cache/tavily_search.jsonl.gz |
321194 |
cache/google_search.jsonl.gz |
1743 |
cache/web_scrape.jsonl.gz |
25 |
tavily_search was saved with cache_expire_hours 0, so those entries do not expire. These files are a snapshot for download. The server reads and appends cache/<tool>.jsonl next to the source tree (see finance_mcp/core/storage/jsonl_cache.py). Writes take an exclusive lock file so several machines can share one JSONL.
gzip -dc cache/tavily_search.jsonl.gz | head -n 1
- Downloads last month
- 214