Every commit scored for change-risk against this repo's own history, so 'elevated' means elevated here rather than on some global curve.
Needs review
59 commits sit in this repo's top risk tercile, which is 33% of the 177scored. The cut is drawn against this codebase's own history rather than a global curve, so a quiet repo still fills its top band, and here it starts at 7.1 out of 10. What pushes a commit up is size and spread together: a large change confined to one area scores below a smaller one scattered across a dozen files.
Commit categories over time, read off the subject line. Fixes carry the accent because that is the series this chart exists to show.
Consistently other-driven across its history.
Ranked by change-risk, highest first. Priority is a tercile of this repo's own distribution, so a quiet repo still fills its top band.
| # | Commit | Author | When | Lines | Risk | Top driver |
|---|---|---|---|---|---|---|
| 1 | da84ef43main works, cli bugs | Edward Sun | 1y ago | +5.6K -49 | 99%Elevated | more lines added than baseline |
| 2 | cc97cb6dchore(release): v0.1.0 – initial public release of TradingAgents | Yijia-Xiao | 1y ago | +5.2K -537 | 99%Elevated | more lines added than baseline |
| 3 | 78ea029aDocker support and Ollama support (#47) | Geeta Chauhan | 1y ago | +2.1K -19 | 99%Elevated | more lines added than baseline |
| 4 | 6abc768cfeat: replace per-agent BM25 memory with persistent append-only decision log | Zhigong Liuclaudeassisted | 3mo ago | +1.1K -353 | 98%Elevated | more lines added than baseline |
| 5 | a6734d71WIP | luohy15 | 10mo ago | +1.3K -1.4K | 98%Elevated | more lines added than baseline |
| 6 | c850ddf1Brief version | Yijia-Xiao | 1y ago | +579 -129 | 97%Elevated | more lines added than baseline |
| 7 | bba14779feat: structured-output Trader and Research Manager (#434, finishes the trio) | Yijia-Xiao | 3mo ago | +519 -69 | 96%Elevated | more lines added than baseline |
| 8 | 0ab323c2Add Alpha Vantage API integration as primary data provider | luohy15claudeagent | 10mo ago | +519 -19 | 96%Elevated | more lines added than baseline |
| 9 | 1ff3f07afix: support commodity/forex/crypto tickers and never invent prices (#781) | Yijia-Xiao | 1mo ago | +460 -53 | 94%Elevated | more lines added than baseline |
| 10 | 0fcf1362feat(agents): rename to sentiment_analyst; integrate StockTwits + Reddit | Yijia-Xiao | 2mo ago | +401 -54 | 94%Elevated | more lines added than baseline |
| 11 | 0fda2451feat: structured-output Portfolio Manager + 5-tier rating consistency (#434) | Yijia-Xiao | 3mo ago | +397 -85 | 94%Elevated | more lines added than baseline |
| 12 | d4dadb82feat: add multi-provider LLM support with thinking configurations | Yijia Xiao | 5mo ago | +647 -966 | 94%Elevated | more lines added than baseline |
| 13 | 79051580feat: add multi-provider LLM support with factory pattern | Yijia Xiao | 5mo ago | +328 -15 | 94%Elevated | more lines added than baseline |
| 14 | c2fa046aTradingAgents-AI | Yijia-Xiao | 1y ago | +460 -0 | 94%Elevated | more lines added than baseline |
| 15 | 22bb91bdfix(llm): structured output for DeepSeek V4 and reasoner | Yijia-Xiao | 2mo ago | +307 -58 | 92%Elevated | more lines added than baseline |
| 16 | 7c37249fchore: release v0.2.4 — structured agents, checkpoint, memory log, providers | Yijia-Xiao | 3mo ago | +562 -2 | 92%Elevated | more lines added than baseline |
| 17 | c07dcf02added fallbacks for tools | Edward Sun | 9mo ago | +218 -20 | 91%Elevated | more lines added than baseline |
| 18 | e80636fcfeat(sentiment): structured output for the Sentiment Analyst | Yijia-Xiao | 1mo ago | +259 -21 | 89%Elevated | more lines added than baseline |
| 19 | d7b40a2afix(graph): resolve instrument identity to stop wrong-company hallucination | Yijia-Xiao | 1mo ago | +390 -44 | 89%Elevated | more lines added than baseline |
| 20 | 9f7abfcbfeat(cli): detect missing provider API keys and persist to .env | Yijia-Xiao | 2mo ago | +261 -24 | 89%Elevated | more lines added than baseline |
| 21 | 7e9e7b83feat: DeepSeek V4 thinking-mode round-trip via DeepSeekChatOpenAI subclass | Yijia-Xiao | 2mo ago | +262 -17 | 89%Elevated | more lines added than baseline |
| 22 | 4cbd4b08feat: add LangGraph checkpoint resume for crash recovery (#594) | Yijia-Xiao | 3mo ago | +349 -21 | 89%Elevated | more lines added than baseline |
| 23 | 2d2c9e6dadd analyst execution planning and timing hooks | CadeYu | 4mo ago | +257 -53 | 89%Elevated | more lines added than baseline |
| 24 | 3543e539fix(dataflows): fall back to Reddit RSS search when JSON 403s | Yijia-Xiao | 1mo ago | +228 -25 | 86%Elevated | more lines added than baseline |
| 25 | aba1880cchore: update .gitignore to official Python template | Yijia Xiao | 5mo ago | +213 -8 | 86%Elevated | more lines added than baseline |
| 26 | 93b87d51fix: analyst status tracking and message deduplication | Yijia Xiao | 5mo ago | +184 -234 | 86%Elevated | more lines added than baseline |
| 27 | 54cdb146feat: add footer statistics tracking with LangChain callbacks | Yijia Xiao | 5mo ago | +277 -112 | 86%Elevated | more lines added than baseline |
| 28 | 47cbb321feat(market): verified market-data snapshot to ground numeric claims | Yijia-Xiao | 1mo ago | +232 -1 | 84%Elevated | more lines added than baseline |
| 29 | 8793336dsync model validation with cli catalog | CadeYu | 4mo ago | +192 -125 | 84%Elevated | more lines added than baseline |
| 30 | 2e67782ffeat(cli): skip interactive LLM selection when configured via environment (#873) | Yijia-Xiao | 1mo ago | +206 -49 | 83%Elevated | more lines added than baseline |
| 31 | 224941d8feat: add post-analysis report saving and fix display truncation | Yijia Xiao | 5mo ago | +157 -174 | 83%Elevated | more lines added than baseline |
| 32 | f10daa28feat(ollama): OLLAMA_BASE_URL end-to-end with endpoint confirmation | Yijia-Xiao | 2mo ago | +230 -8 | 82%Elevated | more lines added than baseline |
| 33 | e1316686fix(llm): MiniMax integration polish vs official docs | Yijia-Xiao | 2mo ago | +172 -14 | 82%Elevated | more lines added than baseline |
| 34 | 78d063dcfeat(reflection): configurable alpha benchmark for non-US tickers | Yijia-Xiao | 2mo ago | +151 -15 | 81%Elevated | more lines added than baseline |
| 35 | d13e9b79feat(config): TRADINGAGENTS_* env-var overlay for DEFAULT_CONFIG | Yijia-Xiao | 2mo ago | +176 -25 | 81%Elevated | more lines added than baseline |
| 36 | e1113880fix: prevent look-ahead bias in backtesting data fetchers (#475) | Yijia-Xiao | 4mo ago | +108 -142 | 81%Elevated | more lines added than baseline |
| 37 | e7ec9800feat: add analysis-only crypto asset mode | CadeYu | 3mo ago | +141 -19 | 79%Elevated | more lines added than baseline |
| 38 | b0f60582feat: add DeepSeek, Qwen, GLM, and Azure OpenAI provider support | Yijia-Xiao | 3mo ago | +163 -61 | 79%Elevated | more lines added than baseline |
| 39 | 7bb2941boptimized yfin fetching to be much faster | Edward Sun | 9mo ago | +130 -10 | 79%Elevated | more lines added than baseline |
| 40 | 2e8a02e8Layout | Yijia-Xiao | 1y ago | +99 -111 | 79%Elevated | more lines added than baseline |
| 41 | 8b7b5b9cLayout | Yijia-Xiao | 1y ago | +93 -115 | 77%Elevated | more lines added than baseline |
| 42 | db9f63faCitations | Yijia-Xiao | 1y ago | +94 -58 | 77%Elevated | more lines added than baseline |
| 43 | 99789f9cAdd support for other backends, such as OpenRouter and olama This aims to offer alternative OpenAI capable api's. This offers people to experiment with running the application locally | maxer137 | 1y ago | +110 -41 | 76%Elevated | more lines added than baseline |
| 44 | 61522e10fix(llm): skip Anthropic effort kwarg on non-supporting models (#831) | Yijia-Xiao | 2mo ago | +106 -2 | 75%Elevated | more lines added than baseline |
| 45 | 08bfe70afix: preserve exchange-qualified tickers across agent prompts | CadeYu | 4mo ago | +91 -24 | 75%Elevated | more lines added than baseline |
| 46 | faaeebacfeat(cli): collapse regional duplicates; refresh Qwen catalog | Yijia-Xiao | 2mo ago | +92 -15 | 74%Elevated | more lines added than baseline |
| 47 | 2c97bad4fix(security): validate ticker before using as path component (#618) | Yijia-Xiao | 2mo ago | +103 -4 | 74%Elevated | more lines added than baseline |
| 48 | 50961b24refactor: rename risky/safe agents to aggressive/conservative | Yijia Xiao | 5mo ago | +99 -99 | 74%Elevated | more lines added than baseline |
| 49 | 10c136f4feat: add Docker support for cross-platform deployment | Yijia-Xiao | 3mo ago | +89 -0 | 72%Elevated | more lines added than baseline |
| 50 | c252bbb1Brief | Yijia-Xiao | 1y ago | +73 -87 | 72%Elevated | more lines added than baseline |
Two views of the same model: where the cuts fall, and what commit shape lands you above them.
Every scored commit, binned on the raw 0 to 10 score rather than the percentile. Percentile ranks are uniform by construction, so that axis has no shape to draw. The dashed lines are the tercile cuts behind each row's priority pill.
The 177 most recent commits, on their own recency sample rather than the feed above: that defaults to risk-sorted, so reusing it would plot only the top tercile and call it the spread. Big and scattered is what the model penalises. Click a dot to open it.
Repowise tracks change history across 102 files in tauricresearch/tradingagents. In the last 90 days 89 files were touched, 361 times in total, most often main.py. Every commit is scored for change risk from its size, spread and the history of the files it touches.
