Skip to main content
ClaudeWave
Skill32.3k repo starsupdated 3d ago

qveris

Paid capability marketplace for global multi-asset data; use it when free Vibe-Trading sources lack coverage, depth, or provider quality, and keep free sources as the default for routine OHLCV.

Install in Claude Code
Copy
git clone --depth 1 https://github.com/HKUDS/Vibe-Trading /tmp/qveris && cp -r /tmp/qveris/agent/src/skills/qveris ~/.claude/skills/qveris
Then start a new Claude Code session; the skill loads automatically.

SKILL.md

# QVeris

QVeris is a paid capability marketplace for global market, fundamental, macro,
derivatives, crypto, China/HK, news, filings, and alternative-data calls. Use it
when the built-in free sources cannot cover the requested dataset, when the user
explicitly asks for QVeris, or when a premium provider is needed for depth such
as options Greeks, analyst/calendar feeds, broad provider comparison, or paid
China/HK/global coverage. For ordinary OHLCV, keep `source: "auto"` on the free
loader chain unless the user explicitly selects `source: "qveris"`.

Signup link: [QVeris via Vibe-Trading](https://qveris.ai/?ref=Vyjjo5G_1cAHJA).
Invite code fallback: `Vyjjo5G_1cAHJA`.

## Workflow

### 1. Discover with `qveris_search`

Call `qveris_search` first. Search is free and returns candidate capabilities
with `tool_id`, provider, parameters, examples, `expected_cost`, billing rule,
categories, and `stats` such as `success_rate` and average execution time.

Good queries are concrete:

- `US listed options chain implied volatility Greeks AAPL`
- `Hong Kong market daily OHLCV financial statements`
- `FRED CPI Treasury yield curve macro series`
- `China A share northbound fund flow daily`

### 2. Inspect with `qveris_inspect`

Call `qveris_inspect` on one or more `tool_id` values before any paid call.
Inspect is free and returns the fuller parameter schema and sample parameters.
Use it to verify required parameters, enum values, date formats, and output
shape before execution.

### 3. Call with `qveris_execute`

Call `qveris_execute` only after the user has enabled QVeris and the tool has
been inspected. Pass the inspected `tool_id`, the exact `parameters`, and a
reasonable `max_response_size`. Use `session_id` and `search_id` when available
so usage can be reconciled later.

## Choosing a Capability

Prefer the candidate with the best combination of:

1. High `stats.success_rate`.
2. Low `expected_cost` for equivalent data.
3. Provider fit for the task: official or specialist sources before generic
   mirrors when the figure matters.
4. Parameter clarity: required fields and examples should match the user's
   requested asset, date range, and granularity.

Do not assume tools with similar names are interchangeable. The measured catalog
has very large quality variance inside the same category; design notes recorded
FMP tools ranging from near-zero to 100% success and some tradefeeds tools at
0%. Surface `success_rate` and `expected_cost` in the answer whenever the user is
choosing among paid options.

## Billing and Modes

- The default free route is Vibe-Trading's built-in public data stack; QVeris
  tools are unavailable in `mode="free"`.
- Discover (`qveris_search`) and inspect (`qveris_inspect`) are free only after
  the user explicitly enables the paid QVeris route.
- Execute (`qveris_execute`) is paid only when the provider call is billable.
- Failed calls, empty results, and parameter errors are not charged by the
  QVeris contract; observed missing-parameter calls returned `success=false`,
  `cost=0.0`, `status=not_charged`, and unchanged balance.
- `mode="free"` means the original free public-data path stays active and
  QVeris is hidden from agent/tool/backtest routing.
- `mode="paid"` permits QVeris discovery, inspect, and execute calls, bounded by
  `budget_credits_per_session`.
- Never hide cost: every successful execute result should preserve `cost` and
  `remaining_credits` for the user.

Configuration can come from `QVERIS_API_KEY` / `QVERIS_BASE_URL`, or from
Settings -> QVeris / `vibe-trading data mode paid`. If QVeris is disabled, in
free mode, or no API key is configured, the tools should be treated as
unavailable.

## Truncation

`qveris_execute` accepts `max_response_size` (`20480` default, `-1` means do not
ask QVeris to truncate). When a response is too large, QVeris returns a result
object containing a message, `truncated_content`, and `full_content_file_url`.
If the full JSON is required, fetch that signed URL and use the full payload.
If a summary is enough, cite that the response was truncated and work only from
`truncated_content`.

## Balance and Usage

Check balance and recent usage through:

- Web: Settings -> QVeris.
- CLI: `vibe-trading data status` or `vibe-trading data usage`.
- Agent/API tools: status is derived from free search/ledger/usage endpoints.

For reconciliation, prefer `execution_id`, `tool_id`, `charge_outcome`, `cost`,
and `remaining_credits` over text summaries.

## References

- [REST API reference](qveris/references/rest-api.md)
- [Coverage map](qveris/references/coverage.md)
vibe-tradingSkill

Professional finance research toolkit — backtesting (10 engines + benchmark comparison panel), factor analysis, Alpha Zoo (462 pre-built alphas across qlib158/alpha101/gtja191/academic/fundamental), options pricing, 90 finance skills, 30 multi-agent swarm teams, Trade Journal analyzer, and Shadow Account (extract → backtest → render) across 25 market-data sources (tushare, yfinance, okx, binance, akshare, baostock, tencent, mootdx, ccxt, futu, mt5, tickerall, local, eastmoney, sina, stooq, yahoo, pykrx, india_broker, qveris, longbridge, plus optional-key finnhub/alphavantage/tiingo/fmp).

adr-hshareSkill

ADR/H-share/A-share cross-listing premium analysis — track pricing gaps between US-listed ADRs, HK-listed H-shares, and A-shares for arbitrage signals, dual-listing valuation, and delisting risk assessment.

akshareSkill

AKShare financial data aggregator (18k+ stars). Free, no API key. Covers A-shares, US, HK, futures, macro, forex. Primary fallback for tushare and yfinance.

alpha-zooSkill

Browse and bench the bundled alpha zoos — prebuilt cross-sectional factor libraries (Kakushadze 101, GTJA 191, Qlib 158, Fama-French / Carhart). Use when the user asks "which alphas exist", wants metadata on a named alpha, or wants to run IC/IR on a whole zoo over a universe.

ashare-pre-st-filterSkill

A 股 ST/*ST 风险预测框架 — 基于最新中报/三季报或业绩预告/快报,预测下一财年是否会因营收、利润、净资产、分红不达标而被风险警示,并将新浪监管处罚记录作为独立证据面纳入风险等级。仅适用于 A 股,不预测财务造假。

asset-allocationSkill

Asset allocation theory and optimizer usage — MPT / Black-Litterman / risk budgeting / all-weather strategy, including guides for 5 optimizers and rebalancing rules.

backtest-diagnoseSkill

Diagnose failed or underperforming backtests, locate the root cause, and fix the issue

behavioral-financeSkill

Behavioral finance applications: theories of overreaction and underreaction, behavioral explanations for momentum and reversal, investor sentiment cycles, cognitive-bias checklists, and debiasing quantitative strategies.