ci-gate

An API contract consumer-impact analyzer that treats every version bump like a seismic event. Go CLI + TypeScript viewer, deterministic reports, CI exit-code gate.

AI run variance and reproducibility analyzer - answer stability, route consistency and timing variance in one deterministic report.

Human-centered frontend quality gate and theory-based UX audit skill with measurable fast-gate checks, weighted scoring, and CI strict mode.

Record and replay AI agent runs offline to catch tool regressions and verify changes to files and databases.

pip install gauntlet-evals · v0.1.0. Merge-blocking evaluation gates for generative AI features: YAML suites run against any HTTP endpoint or Python callable, fail the build on a miss, and emit both a diffable JSON pack and a reviewer document cross-referenced to California's published GenAI risk framework. Aligned to, never approved by.

Java 原生 LLM 可观测与评测：同源 traceId 贯通、国产模型中文评测集、CI 回归门禁。不是 Langfuse 的替代品。

Calibration-first evaluation toolkit for Cloudflare's Clef decision models. Judge cheap, audit confidence.

v0.2.0. Fail-closed evaluation harness for government-facing chat systems: reproducible, provenance-stamped audit verdicts, byte-identical across Python 3.11 to 3.14, with no third-party dependencies. A silent or unreadable target scores zero rather than passing by absence. Two public projects of my own pin it by exact commit.

Julia, standard-library-only diagnostics console that scans polyglot repositories for language-policy violations and architectural anti-patterns, with evidence-backed classifications and scan timing.

Complete security framework for AI Agents, MCP Servers, and RAG pipelines. Features automated MCP fuzzing, cross-agent poisoning detection, persistent memory audits, unified risk scoring, and CI/CD security gates.

Self-hosted accuracy evals + correction loop for document extraction. Bring your own model.

SiteOne Crawler: Crawler/analyzer для SEO, OpenGraph, headings, redirects, 404, performance, accessibility, security.

A CI gate to optimize GitHub stacked PRs (gh stack). Stop rerunning CI on every layer of stacked pull requests.

CLI launch-readiness validator for product, marketing, sales and operations specifications. Scores a spec against a token map and can gate CI.

A Claude Code skill that brands every deliverable by default — and the gate that refuses any file that is off-brand

Field-level evaluation and release gate for LLM document extraction: exact golden set, per-field precision/recall, CI gate that blocks regressions

Prove whether an Arm binary can actually use the chip's matrix hardware. Static ISA verifier + CI gate for AArch64 AI runtimes (SME/SME2, i8mm, dotprod).

Autonomy-readiness audit for repositories you hand to AI agents. 33 deterministic rules across 6 shape-aware dimensions, a ranked remediation roadmap, additive guardrail patches, and a CI gate that fails when an autonomous change makes the repo less safe. 交给 AI Agent 托管前的仓库自治就绪度体检。

Agent and LLM evaluation harness: golden datasets, multi-scorer execution, regression detection across model versions, cost-quality leaderboards, and CI gates for model promotion.

Monday.com → HR warehouse → AI-drafted employment letters in Word and PDF — and the gate that refuses to email one until every fact traces to the warehouse and a human has signed the bytes