Changelog
All notable changes to Daily AI Agent OS.
RSS Feed-
New
Weather Arbitrage Agent — identify market opportunities driven by weather patterns, including commodity impacts, energy demand shifts, and agricultural forecasts. Access via
GET /v1/signals/weather.The Weather Arbitrage Agent ingests real-time meteorological data from NOAA, ECMWF, and proprietary weather models. It correlates severe weather events with historical commodity price movements to generate actionable trading signals. Supports natural gas, wheat, corn, oil, and energy futures. Region filtering available for US, EU, Asia, and global markets. -
New
WebSocket dashboard streaming endpoint (
/ws/dashboard) for real-time signal updates without polling. -
Improved
Python SDK now supports
async/awaitviaAsyncAgentOS. All client methods are available in both sync and async variants. -
Improved
Signal confidence scores now include a breakdown of contributing factors in
metadata.indicators. -
Fixed
Fixed an issue where
min_confidencefilter onGET /v1/signalswas not applied when combined withdirectionfilter.
-
Fixed
Resolved race condition in webhook delivery that could cause duplicate events under high load.
-
Fixed
SDK
client.signals.list()now correctly paginates whenoffsetexceeds total results (returns empty list instead of error). -
Improved
Webhook retry backoff times adjusted — first retry now at 30s (was 60s) for faster delivery recovery.
-
Fixed
Fixed SDK installation error on Python 3.13 due to deprecated
distutilsdependency. -
Improved
Rate limit headers (
X-RateLimit-Remaining,X-RateLimit-Reset) now included in all API responses.
-
New
SEO Content Agent — generate blog posts, product descriptions, and meta tags. Use
POST /v1/agents/seo/run.The SEO agent supports three strategies:blog_postgenerates full articles with headings,product_descriptioncreates conversion-optimized copy, andmeta_tagsgenerates title tags, descriptions, and OG metadata. Configurable tone, word count, and target audience. -
New
Webhook support — receive HTTP notifications for
signal.created,signal.closed, andagent.errorevents. -
Breaking
API key format changed from
dai_*tosk-live-*/sk-test-*. Existing keys continue to work but will be deprecated on April 1, 2026.Generate a new API key from Settings → API Keys in the dashboard. The new format supports test vs. live mode — test keys (sk-test-*) return simulated signals without incurring API charges. Update your environment variables and redeploy. Olddai_*keys will stop working after April 1, 2026. -
Improved
Momentum strategy confidence improved by 12% through incorporation of volume-weighted indicators.
-
New
Sentiment analysis strategy — combine social media sentiment with technical indicators via
strategy="sentiment". -
New
Async agent runs — pass
"async": truetoPOST /v1/agents/{id}/runfor non-blocking execution. Poll with the returnedrun_id. -
Improved
API response times reduced by 40% through query optimization and regional edge caching.
-
Fixed
Fixed timezone handling in
sinceparameter onGET /v1/signals— now correctly parses UTC and offset timestamps.
-
New
Python SDK (
dailyai-agents) released on PyPI with full type hints and IDE autocompletion support. -
New
Mean reversion and breakout strategies added to the Trading Signal Agent.
-
Improved
Dashboard redesigned with real-time signal charts, P&L tracking, and agent run history.
-
New
Initial release of Daily AI Agent OS. Includes the Trading Signal Agent with momentum strategy, REST API, and web dashboard.
Launch features: Trading Signal Agent with momentum strategy, REST API with JWT and API key authentication, signal history and pagination, rate limiting (100 req/min on free tier, 1000 req/min on pro), web dashboard with real-time signal visualization.