#agent-tools

DSH (DeepSeek Harness) plugin that launches and drives a running Godot 4.x game through its in-game TCP interaction server — replaces the godot-mcp MCP server with native agent tools.

Index, search, summarize, and approval-gate local Codex sessions through OpenClaw/MCP. Agent-first recall and dry-run orchestration for Codex-heavy work.

Verified catalog of 300+ APIs and MCP servers with opinions and statistics so that AI agents can call after a one-time API key/OAuth setup. JSON, llms.txt and an MCP server for agents.

Search & read the login-walled Chinese internet for your AI agent — 微信公众号/小红书/抖音/B站 + Twitter/Reddit/HN/YouTube. One CLI, one JSON schema, zero-config WeChat search. 让 AI agent 搜到微信公众号. Install by telling your agent.

Hand a whole browser task off in one call: a decision model drives the page server-side, so a flow costs one call, not a turn per click. Ref-based element tables, code-checked assertions, zero-model macro replay, over the Chrome DevTools Protocol.

Search your Claude Code conversation history by content (not just titles), then resume any session. grep + fzf, no index.

Agent tools from many descriptions in a single integration. Authorize operations, not just tool calls, and approve them when a human is required. The agent gets its own identity and never holds your credentials.

Eyes for AI coding agents 👁️ — render → perceive → report → fix → re-render. A machine-graded visual feedback loop (DOM/contrast/OCR-grounded + optional vision LLM) agents consume to self-correct before claiming done.

Share a full Codex or Claude Code conversation as one link, then import it into a native local thread/session to continue.

Your dashboard is data. Any AI can build it; any human can edit it. A protocol + runtime for agent-composable dashboards — layout-as-data, one guarded control plane, sandboxed agent-authored widgets, MCP server included.

Powerful local-first browser automation for AI agents with real Chromium, isolated profiles, MCP, Maqam approvals, and verifiable evidence.

Agent-friendly CLI for Dida365 / TickTick

Toolkit for keeping Claude Code and Codex instructions, skills, and hooks synchronized.

API workspace in plain markdown. Design specs, send requests, validate contracts from CLI, browser, and coding agents.

Codex and AI-agent Telegram messaging plugin with a local CLI.

Agent-portable SerpBase Google SERP API skill for Codex, Claude Code, opencode, OpenClaw, RAG, SEO and Maps enrichment.

Python SDK + CLI + MCP server for AgentClip — the open-source tool that turns AI agent runs into shareable URLs.

Intent-classification preselector for agent runtimes — decouples library size from token cost

Yandex Search API plug-in for Hermes Agent

agent wayfinding. temporal knowledge store in one sqlite file. no neo4j, no cloud, no lock-in.

Agent-native data workspace: progressive Parquet analysis, durable results and bounded context. Small CLI + MCP. Apache-2.0.

📦 local-first CLI tools for AI agents and humans — calorie tracker, CRM, caffeine half-life, document expiry, dream journal, episode tracker, personal records. pip install clibo.

Deterministic fact-checker for AI coding agents — verify an agent's claims about its own work against the real code, git diff & logs. Local MCP server, cited verdicts, refuses to bluff.

跨AI框架通用的项目可视化后台 · Wiki知识库 · Git版本确认 · 一键开启新对话 · 拖拽布局

Local‑only Query Risk & Scope Analyzer for AI coding prompts — estimate risk, scope, likely files, and token pressure before sending queries to agents like Cursor, Copilot, or Claude.

Fully offline, portable AI agent for Windows/Linux/macOS - llama.cpp launcher with Web UI, OpenAI-compatible API, agent tools, MCP (web search) and reasoning. One-command setup. No cloud, no API keys.

engramx's bi-temporal mistakes + structural code-graph as auto-discoverable Claude Code Skills. Catches recurring mistakes before agents repeat them. Apache 2.0, local SQLite, zero cloud.

Reduce Codex and Claude context waste by measuring MCP tools, repo instructions, and safe token-budget cuts.

🧠 A local, Obsidian-compatible second brain for agents: wikilinked markdown vault, knowledge graph, FTS5 search, live graph web view. CLI + web + MCP. Zero deps.

AI 足迹 MCP Server — 让 AI Agent 管理用户的数字足迹

Universal search, recommendation, compatibility, and intelligence layer for AI agent skills.

Implementation-reuse firewall for AI coding: before an agent adds a new helper/service, surface the existing implementations it overlaps with. Deterministic multi-channel retrieval + semantic adjudication.

Machine-readable preflight for software agents: pass the next unsent tool call; receive ≤2 revision-pinned failure cards or typed UNKNOWN. Offline/zero-install; never authorizes execution.

Instant whole-disk file search for AI agents on Windows - a native MCP server for voidtools Everything. No Python, no es.exe, no SDK DLL.

Tools and skills for research agents: Spare Agent Capacity for Codex and Claude Code, bounded research, and fieldwork templates.

14 PikPak agent skills for AI coding agents (Claude Code / Codex / Cursor / Hermes): onboarding, ad cleanup, dedup, search, WebDAV, speed diagnosis, backup, task watch, Telegram. 只读扫描设计。npx skills add pipiguai87-alt/pikpak-skills

Windows screen control for DeepSeek Harness: screenshot, mouse and keyboard input, UI Automation and window management as agent tools.

Academic research intelligence for AI agents — paper search, scientific literature, citation analysis, arXiv, author metrics, trending research & semantic related-work (pgvector). MCP + x402.

A read-only MCP server exposing Slurm scheduler state to agents — allowlist enforced in code, not requested in a prompt, with progressive disclosure.