A community-driven, multi-agent platform for financial applications
ValueCell exposes 3 tools for news and financial data retrieval. All tools have descriptions and basic input schemas, but quality is inconsistent. Tool names follow verb_noun convention (web_search, get_breaking_news, get_financial_news), which is positive. However, descriptions are surface-level and lack context about when to use each tool vs. alternatives. Parameter descriptions are sparse or missing. Output schemas are not documented anywhere in the provided source. Error handling guidance is absent. The server is built on agno framework with HTTP transport, but the tool definitions lack the depth expected for production agent interactions.
Get breaking news and urgent updates.
Get financial and market news.
Search web for the given query and return a summary of the top results. This function uses the centralized configuration system to create model instances. It supports multiple search providers: - Google (Gemini with search enabled) - when WEB_SEARCH_PROVIDER=google and GOOGLE_API_KEY is set - Perplexity (via OpenRouter) - default fallback
get_breaking_news has empty input schema ({}) with no parameters, but description does not explain what triggers breaking vs. non-breaking news or what data is returned. Missing output schema entirely.
web_search description mentions 'centralized configuration system' and multiple providers (Google, Perplexity) but does not explain to the LLM when to call this tool vs. get_breaking_news or get_financial_news. No selection guidance.
All three tools lack documented output schemas. LLM cannot know what fields to expect, cannot plan downstream tool calls, and cannot extract relevant data.
get_financial_news has optional parameters 'ticker' and 'sector' but descriptions do not explain: Can both be provided? Are they mutually exclusive? What happens if neither is provided? What is returned?
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | D | 57 | <=2025-11-25 | v2 |
| 2026-03-09 | D | 51 | - | v1 |
No error handling guidance visible. Tools may fail (API down, no results, invalid ticker) but no error classification, recovery steps, or actionable error messages documented.
web_search 'query' parameter description is minimal. Baseline: 'average param annotation length: 72 chars'. This is ~40 chars and provides no guidance on query format, special operators, expected length, or failure modes.
No pagination parameters visible (limit, offset, page) for tools that likely return multiple results.