Data freshness verification for AI agents. Probes endpoints for cache staleness, latency, TLS health, and content changes. CLI + MCP server + HTTP API.
-
Updated
Apr 2, 2026 - Go
Data freshness verification for AI agents. Probes endpoints for cache staleness, latency, TLS health, and content changes. CLI + MCP server + HTTP API.
Use API caching in API Gateway to cache your responses. Reduce calls made to backend and improve latency of requests to your API
论衡 — 中文学术/深度长文专用流水线。多 Agent 编排 + 三角验证(文献/数据/案例)+ M 门形式合规 + 实战反馈驱动升级。5000+ 字强推
Declarative, scheduled refresh pipeline for map data: pull GeoJSON/JSON/CSV from HTTP APIs, files or SQL, transform and validate it, detect real changes with an order-stable content hash, and emit a freshness manifest plus a stale-data badge. Includes a composite GitHub Action that rebuilds only when the data moved.
Rego policy that treats data freshness as part of the authorization decision. Escalates on stale data instead of guessing. Fail-closed on invalid input.
A vendor-neutral OpenTelemetry semantic convention for data staleness (data freshness), plus a Python SDK and Go Collector components.
Freshness Kit | TypeScript helpers that label content fresh, aging, or stale using supplied review dates and thresholds.
Public environmental monitoring dashboard for Clear Lake, CA — integrating USGS hydrology, FHABS HAB reports, and OSM geospatial data
Transitive evidence freshness and integrity checks for long-running AI agents, RAG pipelines and autonomous workflows.
Live shadow run of FreshCal on ECB and US Treasury publications, compared with a fixed freshness threshold.
Official public repository for the paper "Data Freshness in Mixed-Memory Intermittently-Powered Systems"
Freshness-aware observability of port environmental monitoring: temporal-graph model, experiments and reproducible results
Validate downloaded partner CSV/Parquet feeds locally after SFTP/S3, then alert on missed rolling receipts.
Business-calendar-aware data freshness checks: schedules, holidays, grace windows, and explainable verdicts for DuckDB, PostgreSQL and dbt.
Claims Registry Kit | Tracks product claims, supporting evidence, and when each claim needs review.
Find stale warehouse tables and trace freshness back to the last run.
Learns each dataset's own arrival distribution instead of applying one fixed freshness rule to everything: 86% fewer false alarms than an "alert if not loaded by 07:00" rule at 99.4% detection, plus lineage-aware alert suppression and per-dashboard trust scores. DuckDB, Python, Airflow-shaped, zero cloud credentials.
To associate your repository with the data-freshness topic, visit your repo's landing page and select "manage topics."