Autonomous AI agents that perform tasks independently
IMPACT Project
Intent-based Multi-agent Platform for Architectural Change Tracking
csinye
Easily create, inspect, and run portable .skill packages for AI agents. Bin: skill.
oyadotai
A plan-don't-react framework for LLM agents. The model emits a typed dataflow plan once; the runtime executes it; the model never reads state it shouldn't.
...moreu1in
🤖 一键部署 Reasonix AI Bot 接入飞书
rsprowler
Drive Claude Code from Telegram: live session cards, policy-gated auto-approval, remote approve/deny, mid-run questions, resume, and a Mini App dashboard.
...moreGitHub Actions
React bindings for @dudko.dev/agent-web: a headless useAgent hook, an AgentProvider context, and optional pre-styled components (chat panel, plan/step view, BYOK key form, WebLLM load bar) that connect the in-browser LLM agent to any React site. UI you ca
...moreGitHub Actions
Lightweight, local-first SQLite event bus for AI agents and developer tools
AdvancingTitans
An evidence-driven Agent engineering quality and control layer for delivery proof, deterministic diagnosis, and bounded asset evolution.
...moreDominick Caponi
Runtime engine for agentic workflows — agent/workflow orchestration with retry, fallback, parallel execution, and short-circuit support.
...morescan5
AI security scanner and runtime enforcement for LLM applications — detects prompt injection, API key leaks, unsafe output, RAG poisoning, and agent hijacking in code and live traffic.
...morerichcanvas
viem-backed ChainProvider adapter for @agenticprimitives/chain-state: readContract-only isRevoked / connection-agnostic signature validation / isAcceptedSessionDelegation reads, pinned to a block for evidence.
...morerichcanvas
Generic relying-party connect client: hand off to the central-auth Home (FedCM-first → OIDC redirect/popup), exchange the code for an SA-signed scoped delegation. The relying half of spec 230/259; no credential ever originates in the relying app.
...morerichcanvas
Vendor-neutral edge ADMISSION layer: the EdgeAdapter interface, trusted-header hygiene, the GatewayAssertion verification port, and the admission pipeline (route resolution + size/depth/envelope + Stage-1 abuse limit). Owns admission only, never authority
...morerichcanvas
Cloudflare adapter for @agenticprimitives/edge-runtime: Request→AdmissionRequest extraction (exact-byte preserving), the Workers rate-limiter binding as a SoftRateLimiter, and Service-Binding dispatch. Imports the platform runtime; the edge core never doe
...morerichcanvas
Generic service-agent capability framework (spec 283): define -> assert -> entitle -> delegate -> primitive-plan -> execute -> receipt. Provider-neutral; all authority evaluators + primitives are injected. Domain catalogs (e.g. @agenticprimitives/treasury
...moreAiOps Enabler
Python SDK for AiOps Enabler — instrument your AI agent's task lifecycle and ratings in 3 lines of code.
Viddi AI <[email protected]>
Metadata, keyword, deep-link, and SEO helper package for Viddi AI (https://viddiai.com), the all-in-one AI creative agent for video, image, and music generation.
...moreablt-ai
Official Python SDK for the TruAgents Unsubscribe API.
Ciel Agent Framework Authors
Ciel Agent Framework - Enterprise multi-agent harness, model-agnostic, deploy-agnostic
planet-libre
React UI library for chat agents: headless primitives, themeable styled components, OpenAI-compatible SSE runtime, and a typed tool-call widget system.
...moreUncleSaurus
Observability feature for Kestrel Sovereign — hook-driven lifecycle logging, per-agent LLM-call views, and a fleet swimlane UI (agent-tree selector + lanes/sublanes), with a shared Prometheus registry
...morePiotr Konowrocki
CLI for browsing Allegro offers, managing cart, and tracking packages — human-readable and LLM-agent friendly
AI Runtime Supervisor Team
Control plane for production AI-agent work: plan, govern, and verify agent runs.
Tabstack
Tabstack tools for LangChain. Schema-enforced web extraction and research, backed by the official Tabstack SDK.