Fetching metadata from the HF Docker repository... 16 days ago
-
fix: identify invalid evidence pointers for correction
yf176
-
fix: preserve failed agent token usage in trace
yf176
23 days ago
-
fix: validate macro evidence within structured correction
yf176
-
fix: retry transient OpenRouter gateway failures
yf176
24 days ago
-
fix: remove unavailable Finnhub candle path
yf176
25 days ago
-
fix: distinguish Finnhub candle access denial
yf176
-
fix: use Yahoo daily prices before Finnhub candles
yf176
26 days ago
-
fix: expose Finnhub candle denial reason in smoke test
yf176
-
fix: prefer dated MarketData options chains
yf176
-
fix(smoke): show elapsed provider time for coverage
yf176
-
fix(smoke): report effective provider coverage
yf176
-
fix(options): preserve Yahoo auth errors and add dated chain fallback
yf176
-
feat: add UI provider smoke test
yf176
27 days ago
-
fix(options): expose empty Yahoo response diagnostics
yf176
-
fix(quant): keep asset citations within frozen inputs
yf176
29 days ago
-
fix(options): retry empty Yahoo expiration response
yf176
-
fix(llm): retry OpenRouter idle timeout once
yf176
30 days ago
-
fix(llm): handle OpenRouter error payloads
yf176
-
fix(trend): reduce prompt and improve JSON retry
yf176
-
fix(llm): use compatible OpenRouter token limit
yf176
-
fix(llm): enforce OpenRouter structured output
yf176
-
fix(llm): handle empty OpenRouter completions
yf176
-
fix(llm): honor configured OpenRouter model
yf176
-
fix: surface configured LLM timeouts
yf176
-
fix: redact Finnhub provider errors
yf176
-
fix: keep Gemini credentials out of request URLs
yf176
-
fix: harden live structured agent output
yf176
-
fix: preserve yfinance fallback without Finnhub key
yf176
-
fix: redact provider credentials from runtime errors
yf176
-
feat: integrate governed quant support contexts
yf176
-
feat: integrate governed quant runtime
yf176
-
feat: add typed VizAgent execution
yf176
-
refactor: share typed DataFetcher evidence resolution
yf176
-
feat: verify typed evidence references
yf176
-
feat: add typed OptionsFlowWatcher execution
yf176
-
feat: add typed DataFetcher execution
yf176
-
feat: normalize provider data for typed runs
yf176
-
feat: resolve published quant inputs
yf176
-
feat: persist typed quant run records
yf176
-
feat: enforce quant ACL during RAG retrieval
yf176
-
feat: assemble specialized analyst outputs
yf176
-
feat: publish typed artifacts atomically
yf176
-
feat: enforce scoped quant artifact access
yf176
-
feat: validate structured LLM output and bind typed templates
yf176
-
Complete V7.1 replay and evidence guardrails
yf176
-
Validate partial OHLCV provider rows
yf176
-
Use Finnhub candles before yfinance fallback
yf176
-
Implement V7 session-delta committee tracking
yf176
-
feat: wire options context into v6 watcher
yf176
-
feat: V4.5 — frictionless session init + OpenRouter support
yf176
- Ticker-first auto-init: _extract_tickers_from_text(), _generate_goal(),
_auto_init() in slack.py — no /poly init required (FR-1, FR-2, FR-3, FR-4)
- Friendly no-session prompt when no tickers detected (FR-1.4)
- /poly init --force + existing session warning (FR-5.2/5.3)
- /poly goal + @coordinator goal mention handler (FR-3.3, FR-7.1)
- COMMAND_REGISTRY updated with --force and goal entries (FR-7.1)
- OpenRouterAdapter in llm_service.py — OpenAI-compatible, usage.cost passthrough,
ConfigurationError on 400 (FR-OR-1 through FR-OR-7)
- _PROVIDER_MAP refactor covers anthropic/openai/gemini/openrouter
- version bumped to 4.5.0 in __init__.py and pyproject.toml"