Search

Repair interrupted AI agent tool-call histories into safe, replayable transcripts.

Prompton DocsAgent for conversational Starlight docs

A local, read-only dashboard showing what a running Claude Code session has in its context window: observed token usage, growth trajectory, composition, evidence, tools, and clues.

pi extension: the dsh-agent-browser tool surface (snapshot+ref browser automation over agent-browser) for pi sessions. Viewing via agent-browser's own dashboard URL.

No description available

No description available

Official Python SDK for MarkanM Chat — build bots & AI agents with 13 AI providers (Sarvam, Groq, Gemini, OpenAI, Anthropic, Ollama & more)

Create local AGENTS.md and CLAUDE.md aliases in Git repositories

promptex 的 Python SDK 最小原型（alpha）：宣告 skill 並投影成 SKILL.md

Standalone Playwright-first autonomous browser agent for SERA — provider-agnostic LLM planner, real-view human-takeoverable browser, CDP AX tree, multi-tab, stealth, history/replay, structured extraction.

Agentic Pay for Vendure — x402 stablecoin payment method for Vendure letting AI shopping agents pay via the x402 protocol.

DeepSeek Harness plugin: agent registry with layered prompts, parallel delegation, Feishu bots and token usage

BobSentry Python SDK for runtime authorization of consequential AI-agent actions

Guardrails for LLM agents: conditional constitution scoring, PII redaction, crisis detection.

Agent integrations for CapSolver — framework-agnostic tool definitions + LangChain tools.

AI model: primitive-ai/Ornith-1.5-35B-A3B-agentic-NVFP4-FP8

AI model: mradermacher/AgenticOCR-8B-GGUF

Your Personal AI super intelligence. A brain that builds a local-first memory of your life, a fantastic orchestrator of agent fleets and workflows, and a deep researcher.

The Fastest AI Memory Model - Your Second Brain

A small, provider-neutral agent harness written in Rust