320 fully free open-source tools with no paid tier, curated: pipeline-diff, Clai, OpenDots (CopilotKit). Every entry lists license, pricing, and links.
pipeline-diff· tool, 2026-10-05 pipeline-diff is an open-source pipeline change tracker for HubSpot and Salesforce: it rebuilds what changed in your sales pipeline over any window (deals created, won, lost, pushed out, amount and forecast-category moves) from the CRM's own change history, onto a self-hosted waterfall dashboard and agent-readable files. Maker-positioned as an OSS alternative to Clari/Gong's pipeline waterfall. MIT; self-hostable (pip install, read-only CRM). Free. Published 2026-09-29; nascent (~1 star).
Clai· tool, 2026-10-03 Clai is an open-source command-line AI harness for terminal power users: context-engineering and agent workflows you run as a standalone CLI or embed as a Go package. Open source: yes (MIT); self-hostable. Free OSS; ~145 stars; established and active (~925 commits since 2024).
OpenDots (CopilotKit)· tool, 2026-10-03 OpenDots is CopilotKit's open-source template for persistent, always-on AI agents that each get their own computer and move between text, calls, and Slack, on web and mobile. Open source: yes (MIT); self-hostable; bring your own OpenAI-compatible model. Free OSS; ~2.2k stars; new (~18 commits). Distinct from the separately-listed anil-matcha/open-dots.
Strata· tool, 2026-10-03 Strata is an open-source local inference engine that runs large open models (the maker demos a 125B-parameter model) on a single consumer GPU, exposing an OpenAI/Anthropic-compatible API on localhost. Open source: yes (MIT); self-hostable. Free OSS; ~8.2k stars; active (~628 commits). Needs an NVIDIA or AMD GPU with 12 GB or more.
Dataverse Business Skills· tool, 2026-10-02 Dataverse Business Skills is Microsoft's open (MIT) catalog of agent skills for Microsoft Dataverse: natural-language instructions that capture CRM and Dynamics 365 Sales processes (logging activity, prepping briefings, outreach, forecasting) as files an agent reads and follows. Open source: yes (MIT) for the skill definitions; NOT standalone — requires a Microsoft Dataverse / Dynamics 365 environment. Free (MIT); ~52 stars; from microsoft.
Microsoft Agent Framework· tool, 2026-10-02 Microsoft Agent Framework is an open (MIT) framework for building, orchestrating, and deploying AI agents and multi-agent workflows, with first-class support for both Python and .NET. Open source: yes (MIT); self-hostable. Free OSS; ~13.9k stars; very active (~3,400+ commits) from microsoft.
Nanobrowser· tool, 2026-10-02 Nanobrowser is an open-source Chrome extension for AI-powered web automation: it runs multi-agent workflows inside your own browser using your own LLM API key, positioned as an alternative to OpenAI Operator. Open source: yes (Apache-2.0); self-hostable (runs in your browser, bring your own key). Free OSS; ~13.9k stars; active (~376 commits).
Kavla· tool, 2026-10-01 Kavla is a self-hostable infinite-canvas data workspace: it pairs a tldraw canvas with an in-browser DuckDB and an OpenAI-compatible agent (works with the Codex CLI), so you can lay out, query, and explore data spatially and let an agent drive the analysis. Open source: yes (Apache-2.0); self-hostable. Free OSS; early project (~19 stars; ~75 commits).
Magnitude· tool, 2026-10-01 Magnitude is an open-source inference engine for agents that optimizes itself for your exact hardware — it compiles and tunes its own kernels on your device, so open models run (per the maker's benchmark) up to 2x faster than llama.cpp, on Apple Silicon, NVIDIA, AMD, or CPU. Open source: yes (Apache-2.0); self-hostable. Free OSS; ~6k stars; very active (~1,000+ commits).
RRSI· tool, 2026-10-01 RRSI (Regularized Recursive Self-Improvement) is a Google Research framework that lets an LLM agent evolve its own harness — prompts, control flow, tools, memory, skills, sub-agents — around a frozen model, while a regularized search curbs the overfitting that plain harness-evolution produces on the training tasks. Open source: yes (Apache-2.0); self-hostable. Free OSS; ~1.1k stars; paper-backed (arXiv 2609.24972).
Mingbird· tool, 2026-09-30 Mingbird is a local-first agent harness that gets small 2–9B models to finish real multi-step tasks on an ordinary laptop, running offline against a local Ollama backend. Open source: yes (Apache-2.0); self-hostable (local Ollama; one-click zero-outbound mode; config in ~/.ollama_agent/config.json). Free OSS; ~1 star (brand-new, created 2026-09-17); active (30 commits, v1.8.2, 461-test suite, CI on Linux/macOS).
Open Dots· tool, 2026-09-30 Open Dots is a self-hosted AI agent workspace — chat, tool integration, and optional computer automation — with a deny-by-default gateway that pauses higher-risk actions for explicit human approval before they run. Open source: yes (MIT); self-hostable. Free OSS; ~4.8k stars; active (104 commits). An OSS alternative to hosted 'Dots'-style agent workspaces.
Jeeves· tool, 2026-09-29 Jeeves is a reasoning-augmented Qwen3.5-9B decision classifier that produces calibrated yes/no/rating calls and exposes the reasoning behind each one. Open source: yes (MIT); self-hostable (download weights, local CUDA inference or Python SDK). Free; ~211 stars; maker PostHog.
Jeff· tool, 2026-09-29 Jeff is a family of small (0.8B–2B) zero-shot decision and classification models you embed directly in code for routing, moderation, and intent calls at ~22–28ms latency. Open source: yes (MIT code + Apache-2.0 model weights); self-hostable (local jeff-serve on CUDA or Apple MLX, weights pulled from HuggingFace, no external API). Free; ~952 stars; maker firelex.
Jitsu· tool, 2026-09-29 Jitsu is an open-source alternative to Segment: it collects event data from sites, apps, and servers and routes it in real time to data warehouses and third-party tools (PostHog, GA4, Meta, Ads). Open source: yes (MIT); self-hostable, with no feature gating; hosted option at jitsu.com. Free (open-core); ~5.1k stars; maker jitsucom.
OpenShell· tool, 2026-09-29 OpenShell is a kernel-level runtime sandbox that lets fleets of autonomous AI agents use files, packages, APIs, and credentials while enforcing security policy without granting unrestricted system access. Open source: yes (Apache-2.0); self-hostable (kernel-level runtime you run yourself). Free; ~10.3k stars; maker NVIDIA; supports Claude Code.
Sensored· tool, 2026-09-29 Sensored is a streaming-first PII/PHI redactor that sanitizes text before it reaches an LLM. Open source: yes (MIT); self-hostable (core regex + NER runs locally; an optional AI-confirmation step calls an external API and fails open). Free; ~1 star; maker atomicpages.
textsnap· tool, 2026-09-29 textsnap converts any image, screenshot, or webpage into plaintext locally via OCR, running fully offline after the first model download. Open source: yes (MIT project; models Apache-2.0); self-hostable (pip install, ONNX runtime, no GPU, no cloud, no API keys). Free; ~186 stars; maker kouhxp.
Charter· tool, 2026-09-28 Charter is a Python library that lets you declare agent API tools as Pydantic schemas instead of hand-implementing them, auto-handling request construction, auth, wire format, and response trimming. Open source: yes (Apache-2.0); self-hostable; pricing free. In-process, no telemetry; launched 2026-09-28; 1 star.
Hindsight· tool, 2026-09-28 Hindsight is an agent memory system exposing retain/recall/reflect over four memory networks — world, experience, opinion, observation — with per-bank MCP endpoints and 60+ integrations. Open source: yes (MIT); self-hostable (docker run; PostgreSQL/Oracle AI DB). Free OSS (optional hosted from vectorize.io); ~39.9k stars; active.
Kern Sandbox· tool, 2026-09-28 Kern Sandbox is a rootless, daemonless container runtime that runs LLM-generated code and agent tool calls in kernel-enforced isolation, callable from a CLI or Python/Node SDKs with MCP support. Open source: yes (Apache-2.0); self-hostable; pricing free. Single static binary; 427 stars; active.
Open-Agent-DB· tool, 2026-09-28 Open-Agent-DB is an offline-first catalog and search engine for AI agent skills, MCP servers, and project directives, queryable via `npx open-agent-db search` with full-text (SQLite FTS5) and dense-vector semantic search. Open source: yes (MIT); self-hostable; pricing free. Local SQLite, offline-first; launched 2026-09-28.
OpenAPPA· tool, 2026-09-28 OpenAPPA is a deterministic guardrail layer that intercepts agent tool calls and verifies data flows against declarative policies, blocking data exfiltration from prompt injection or hallucination without breaking agent function. Open source: yes (MIT); self-hostable; pricing free. Runs in-process; 26 stars; active.
Subscrio· tool, 2026-09-28 Subscrio is an open-source entitlement engine for .NET and TypeScript that models subscription plans, feature access, limits, and per-customer overrides in your own database. Open source: yes (MIT); self-hostable; pricing free. Runs as an in-app library, no hosted service; launched 2026-09-28; pre-1.0.
Apache Airflow· tool, 2026-09-28 Apache Airflow is open source (Apache-2.0): a platform to programmatically author, schedule and monitor workflows as directed acyclic graphs (DAGs) written in Python. Pipelines are maintainable, versionable and testable; a rich web UI visualizes runs and troubleshooting, with extensible operators and Jinja templating. An Apache Software Foundation project.
Flowable· tool, 2026-09-28 Flowable is a compact, open-source (Apache-2.0) BPM platform for developers: a fast, tried-and-tested BPMN process engine, CMMN case engine, and DMN rule engine written in Java. Runs embedded in a JVM app or as a server/cluster/cloud service. Apache-2.0 open source with a committed community edition; full documentation at flowable.org.
Freescout· tool, 2026-09-28 FreeScout is a lightweight, free and open source (AGPL-3.0) help desk and shared inbox built in PHP/Laravel — a self-hosted, OSS open source alternative to Zendesk (and Help Scout) with unlimited agents, tickets and mailboxes, plus email, live chat and messaging channels.
Office365-REST-Python-Client· tool, 2026-09-28 MIT-licensed Python client for SharePoint REST and Microsoft Graph — OAuth2 via app registration with client certificate or secret. The most complete community Python option for SharePoint, OneDrive and Microsoft 365 file operations, working within Microsoft's terms; complements the official msgraph-sdk-python.
OpenSERP· tool, 2026-09-28 Self-hosted, open-source SERP API and CLI (MIT) that returns structured search results from Google, Bing, Yandex, Baidu, DuckDuckGo, and Ecosia via a real rendering browser — no API keys or per-search fees. Runs as a localhost HTTP API or a terminal command, with JSON, Markdown, plain-text, and NdJSON output. An open alternative to paid SERP APIs.
SerpBear· tool, 2026-09-28 SerpBear is free, open-source rank tracker software (MIT, self-hosted): monitor daily Google keyword rankings for unlimited domains and keywords, with a Google Search Console integration for real impressions/visits and a Google keyword ranking API for reporting tools. Keyword research uses Google Ads search-volume data. No per-keyword SaaS fee.
SurfSense· tool, 2026-09-28 SurfSense is an air-gapped, open-source (Apache-2.0) NotebookLM alternative: an AI research app that parses, indexes and embeds your documents entirely on your own machine (local SQLite, no telemetry), then generates summaries, reports, slides, podcasts and study guides. SurfSense AI runs on local models or any OpenAI-compatible API, and ships an MCP server for connecting external sources.
HydraDB· tool, 2026-09-28 Open-source distributed graph database in Rust built on object storage — decouples storage from compute so data nodes and indexers scale independently. Neo4j-compatible via Bolt 5.x, with an HTTP query API supporting OpenCypher and GraphBLAS traversal. An infra / agent-memory / context layer.
GeoLift· tool, 2026-09-28 Meta's open-source geo-experiment framework (MIT) that uses Synthetic Control Methods to measure the true incremental lift of ad campaigns when platform-reported conversions aren't enough — geo lift testing that designs geo tests and estimates the causal effect. An R package from facebookincubator.
browser-session-ctl· tool, 2026-09-27 browser-session-ctl is a Chrome extension plus local sidecar that lets shell commands click, type in, and read your existing logged-in browser session, for use with Claude. Open source: yes (GPL-3.0-or-later); self-hostable; pricing free. Localhost-only, JavaScript, 27 commits.
Chimera· tool, 2026-09-27 Chimera is a local-first workspace where AI agents propose actions classified auto, confirm, or blocked, with every decision written to a tamper-evident audit chain. Open source: yes (Apache-2.0); self-hostable; pricing free. Model-independent (Codex, Claude Code, Bedrock, OpenRouter); source beta.
Provenance Gate· tool, 2026-09-27 Provenance Gate is a deterministic authorization gateway that tracks where each tool-call value originated so an AI agent cannot act on untrusted data in side-effect operations. Open source: yes (Apache-2.0); self-hostable; pricing free. Early PoC; runs as a Claude Code plugin.
swe-mux· tool, 2026-09-27 swe-mux is a terminal multiplexer that runs multiple coding agents (Claude Code, Codex) side by side with live per-session status, reachable from a phone over Tailscale. Open source: yes (Apache-2.0); self-hostable; pricing free. Python 3.12+, PyPI wheel, 6 stars, fresh.
duckdb.mk· tool, 2026-09-26 duckdb.mk is a minimal Make-based build system that parses SQL file dependencies to build reproducible DuckDB and Parquet data pipelines, dbt-style, with no config. Open source: yes (MIT); self-hostable, needs DuckDB 1.4.1+, GNU Make and curl. Brand new and unproven: 0 stars, created 2026-09-25.
Keel· tool, 2026-09-26 Keel is a gated Rust harness that runs AI coding-agent output through five sequential verification gates and emits cryptographically-linked evidence bundles requiring human sign-off before deploy. Open source: yes (Apache-2.0 per GitHub SPDX); self-hostable; pricing free. 2 stars, fresh (last push 2026-09-26) and unproven.
proxy-scraper· tool, 2026-09-26 proxy-scraper is a self-hostable proxy scraper and verifier that pulls candidate proxies from 700+ sources, then runs honeypot, injected-script and TLS checks, and can serve a rotating proxy plus an MCP server. Open source: yes (MIT); self-hostable via pipx, pip or Docker. 5 stars, created 2026-09-24.
browser-use web-ui· tool, 2026-09-26 A Gradio-based web interface for running browser-use AI agents in your browser without the command line. Supports most browser-use functionality, custom browser and persistent sessions, and multiple LLM backends — Google, OpenAI, Azure OpenAI, Anthropic, DeepSeek, and Ollama. MIT-licensed.
One API· tool, 2026-09-26 Self-hostable LLM API gateway (MIT) that unifies 30+ providers — OpenAI, Azure, Anthropic Claude, Gemini, DeepSeek, and major Chinese models — behind a single OpenAI-compatible endpoint. Manages and redistributes provider keys so clients just swap the base URL, and centralizes access control, quotas, and usage monitoring. Ships as a single binary or Docker; an open alternative to LiteLLM.
taskt· tool, 2026-09-26 Free, open-source robotic process automation (RPA) built in C#/.NET for Windows: a visual, no-code bot designer with element and screen recording, dozens of automation commands, desktop and web app automation, and Excel integration. Optionally runs custom .NET, PowerShell, or VBScript for advanced steps. Apache-2.0.
ToolHive Registry Catalog· tool, 2026-09-26 Open-source registry catalog of MCP servers from Stacklok (Apache-2.0). Demonstrates twin catalog streams — servers for executable components and skills for reusable prompt workflows — with an allowedTools array that bounds which tools a skill may invoke, keeping agent execution scoped.
Cold Outbound Skills· tool, 2026-09-26 Thirty open-source Claude Code skills covering the cold-email motion end to end — lead sourcing, copywriting, deliverability/infrastructure setup and campaign operations — packaged by growth operator Eric Nowoslawski. MIT-licensed; distinct from the Corey Haines marketingskills collection.
Ctxfw· tool, 2026-09-25 Ctxfw is an Apache-2.0, in-memory Tree-Sitter AST 'context firewall' — an MCP server that prunes what a coding agent sends to the model, trimming the code context down before it costs tokens. It is on PyPI and self-hostable. It is nascent (2 stars). Its headline '-72.4% token mass' figure is a self-reported, unbenchmarked project claim, not a measured result.
oh-my-subagents· tool, 2026-09-25 oh-my-subagents is local subagent orchestration for Codex and Claude — persistent task state, reusable teams, and interruption recovery. Its licensing is split, not simply 'open source': the core orchestration code is MIT, but the Console UI is source-available under the n8n-derived Sustainable Use License v1.0 (internal/non-commercial use only; no commercial redistribution), which is not OSI-open. It has 142 stars.
Sifthound· tool, 2026-09-25 Sifthound is an MIT-licensed, self-hosted web search / extract / crawl / map API for AI agents, exposing a Tavily-compatible interface plus an MCP server. It is brand-new and unproven — created 2026-09-23, last pushed 2026-09-25, 0 stars — so treat it as a fresh drop, not an adopted tool. The value is that it is self-hostable: an owned endpoint you run rather than a paid search API you call.
Skillpack· tool, 2026-09-25 Skillpack is an MIT-licensed, self-hostable, multi-tenant hub to share, version and publish agent skills across MCP clients — an OAuth 2.1 MCP server with a native CLI and a web UI. Made by Stan Girard (of Quivr), authorship confirmed. It is both a hub-style service and a real self-hostable OSS core. It is very new: 5 stars, created 2026-09-14, last pushed 2026-09-25.
Zro CLI· tool, 2026-09-25 Zro CLI is an MIT-licensed command-line tool that runs your existing agent harnesses — Claude Code, Codex, OpenCode, Hermes — against open models, so you can point the coding harness you already use at an open-weight backend. It is self-hostable and early: 2 stars, near-zero traction, created 2026-07-22, last pushed 2026-09-24.
AgentRun· tool, 2026-09-24 AgentRun is a workflow DSL from Parcha-ai — expressed in JSON or TypeScript — for turning repeatable AI-agent work into orchestrated, typed workflows. It ships beta npm packages and is self-hostable under Apache-2.0. It is brand-new: created 2026-09-23, only 20 stars, and explicitly beta, so treat it as an early bet rather than a settled orchestration layer.
Beacon· tool, 2026-09-24 Beacon is an MIT-licensed, local-first memory layer that captures AI-coding-agent sessions across harnesses — Claude Code, Cursor, Codex, OpenCode and others — and serves reusable knowledge over MCP or Agent Skills. It has 1,493 stars, is actively releasing (v1.3.x this week), and self-hosts. Its claim to grade runs into reusable skills is the tool's own, not a verified result.
BoxHaven· tool, 2026-09-24 BoxHaven provides persistent remote dev machines — cloud VMs — for AI agents such as Claude Code and Codex, and is self-hostable. It is licensed AGPL-3.0, a strong/network copyleft license: network use of a modified version triggers source-sharing obligations, which matters for anyone running it as a hosted or SaaS offering. It is nascent (6 stars).
jev-seo· tool, 2026-09-24 jev-seo is an MIT-licensed Rust CLI plus MCP server that runs roughly 50-rule SEO/GEO audits, performs live crawls, computes GEO scores, detects rank drift and enforces CI gates. It is free and self-hostable, and because it exposes an MCP server it can also be driven by agents. It positions as an OSS alternative to paid SEO/GEO auditors with direct GTM utility.
Opaline· tool, 2026-09-24 Opaline is an MIT-licensed CLI that gives PostHog-style session analytics — token cost and usage — for Claude Code and Codex, and is self-hostable. It is a mature project (created February 2026, 298 stars), not a brand-new drop; the recent event is its rename and relaunch from rudel to Opaline. It has been maturing since February.
Repairo· tool, 2026-09-24 Repairo is an Apache-2.0 CLI that detects breaking OpenAPI/API drift and auto-generates AST-transformed, compile-checked fix PRs in TypeScript, Python or Go. It ships an npm package (repairo-cli), CI and a hosted beta, with 184 commits behind it. It carries only 2 stars, so treat it as early and unproven — but it is a real project, not a toy.
Whiteboard· tool, 2026-09-24 Whiteboard is an MIT-licensed desktop canvas where humans and coding agents such as Claude Code collaborate on software design, with AST-aware semantic diffs and decision logging. It is actively shipping — 434 commits, 303 stars, created 2026-08-18, last pushed 2026-09-24. It self-hosts as a desktop app and serves as a shared design surface between developers and their agents.
AWS TOLAP· tool, 2026-09-23 Apache-2.0 protocol from AWS Labs — the Tool-Object Level Access Protocol — that enforces object-level access control (column, row, field) around the tool/function layer so data-access policies travel with the data rather than the agent. It is protocol-agnostic across MCP, LangChain, Bedrock Agents and Semantic Kernel, and ships enforcement SDKs for .NET, Python and TypeScript plus a reference policy server. A legitimate first-party AWS launch, still almost undiscovered (5 stars).
Basta· tool, 2026-09-23 MIT-licensed dead-code detector with a Rust engine that finds unused files, exports, imports and declarations across JavaScript/TypeScript and Python. It is a command-line developer tool for codebase hygiene — you run it against a repo and it reports what is provably unreferenced so it can be safely removed. Rust is the analysis engine, not a target language; the languages it analyzes are JS/TS and Python. Actively maintained (npm v0.3.0 published 2026-09-22).
Tenderness· tool, 2026-09-23 Apache-2.0 Python library that renders synthetic, deterministic documents from text and images via Cairo/Pango to produce labeled datasets for training vision-language and OCR models. It is a rendering library, not a hosted generator app — you call it in code to emit page images with known ground-truth, so the data-prep layer for a document-extraction model is reproducible instead of hand-collected. Early and small (7 stars), for teams needing controlled training data, not a turnkey product.
Buzz· tool, 2026-09-22 Apache-2.0, self-hostable workspace from Block where humans and AI agents collaborate in shared rooms and channels built on a Nostr relay you run yourself. Agents hold their own scoped keys and every event is logged, so a team can hand agents real work — code review, bug triage, workflow orchestration — as permission-controlled, auditable teammates on infrastructure it owns rather than a vendor's.
Drop· tool, 2026-09-22 Apache-2.0 rootless Linux sandbox that runs untrusted programs and agent-generated code in disposable, isolated home directories using user namespaces, with optional gVisor for a stronger boundary. Access is governed by a TOML file and nothing needs root or a container runtime, so you can execute a coding agent's output or a third-party dependency without giving it your real user account.
Strands harness-sdk· tool, 2026-09-22 Apache-2.0 SDK (Python and TypeScript) from the AWS-backed Strands Agents project for building AI-agent harnesses that run in-process, with no hosted control plane. It is model-agnostic — first-class Amazon Bedrock plus Anthropic, OpenAI and Gemini, and local models via Ollama — so an agent you build stays portable across providers and can run entirely on your own hardware instead of being tied to one vendor's runtime.
Google AX· tool, 2026-09-21 Apache-2.0 agentic orchestrator from Google (Go) for running autonomous agent workloads at scale. You declare an agent task in a YAML manifest, its workspace and its gateway, and AX sandboxes the task, wires up the workspace, fences its network, and runs it, on Kubernetes, through a control plane. It integrates with LLM providers you supply, so it handles the isolation and scale-out while you bring the model. Early observers describe it as Kubernetes for agent executions.
motif· tool, 2026-09-21 Apache-2.0 shared-memory system for teams of AI coding agents (TypeScript) that collects agent sessions from Claude Code, Codex and Cursor into one queryable graph of decisions, files and the ties between them, so a team can recall why something was done and stay consistent across tools instead of every session being an island. Self-hostable via Docker or npm with no license key, no feature flags and no limits on members, sessions or projects.
ZCode· tool, 2026-09-21 Apache-2.0 agentic development environment from Z.ai (Zhipu): a TypeScript Electron app that puts the agent conversation at the center, surrounded by a file manager, terminal, Git panel and live browser preview, with desktop, browser and terminal interfaces and an Agent CLI runtime. Z.ai open-sourced the full stack (client, backend, UI, Agent CLI). It is built as the official harness for GLM models and does not ship Claude or GPT support, though the license lets you add integrations.
Cloudflare security-audit-skill· tool, 2026-09-20 MIT coding-agent skill from Cloudflare (JavaScript) that turns an agent into a multi-phase security auditor, built on the assumption the agent will hallucinate, so it makes the agent prove everything. Recon maps the app's trust boundaries, isolated hunter agents attack from separate angles (injection, access control, business logic, crypto), validation agents try to disprove each finding, and the agent that checks a finding is never the one that found it. Output is verified and machine-readable.
Agent Governance Toolkit· tool, 2026-09-19 MIT public-preview governance stack from Microsoft for putting deterministic controls around autonomous agents. It intercepts tool calls, messages and delegations before execution, evaluates YAML, OPA or Cedar policies, assigns agent identity, records audit decisions, and adds sandboxing, kill switches and MCP security. The Python stack installs locally, with SDKs or adapters for TypeScript, .NET, Rust, Go, OpenAI Agents SDK, LangGraph, CrewAI, AutoGen, Semantic Kernel and other runtimes.
Laya· tool, 2026-09-19 Apache-2.0 local decision engine (Python, open weights) for classifying and scoring text without generating prose. It returns typed choices, ordinal scores and boolean probabilities in one forward pass, with presets for routing, guardrails, moderation and support triage. Three checkpoints cover English, multilingual and task-tuned work. The project reports 33 ms on a T4, but also discloses weak zero-shot performance on the base model and the need for domain calibration.
Bastionskill· tool, 2026-09-18 MIT skill-poisoning scanner (Python, by Stefano Rizzello) that reads an agent skill before you install it and flags malicious bundled code: hidden network egress, secret theft, hook-install persistence, destructive commands, obfuscation, opaque binaries, and capabilities the SKILL.md never declared. Static analysis only, the skill's code is never run; installs via pip with zero dependencies (3.10+); scans a local skill, a directory, or a remote owner/repo. v0.2.0 returns allow / review / block.
Octop· tool, 2026-09-18 MIT self-hosted AI assistant platform from Tencent Cloud (Python) that runs multi-user, multi-agent for a household or small team from a single instance: one process serves a web dashboard, a CLI, chat channels (Feishu, DingTalk, QQ, Discord, WeCom), an HTTP/SSE/WebSocket API, IDE integration over ACP, and cron automation, all sharing the same agents and data on your own machine. You bring your own LLM provider (OpenAI, DashScope, Ollama and others); nothing core requires a paid Tencent service.
skillbox· tool, 2026-09-18 MIT self-hosted store for AI-agent skills (TypeScript, by Kitze) that keeps your scattered markdown skills versioned in one place and serves them over MCP with client-key authentication, so any agent can pull the skill it needs instead of you copy-pasting instructions between projects. Runs via Docker with no mandatory paid dependency; optional skill-recommendation features can use a paid AI provider but are not required for the core.
Talos· tool, 2026-09-18 MIT self-hosted agent for the terminal and Telegram (Python) built around a deterministic security kernel: before any tool call runs, PolicyKernel.decide() returns ALLOW, NEEDS_HUMAN or DENY, and an approval is a single-use token bound to the exact arguments and valid for 30 seconds. The design principle is 'the model proposes, it never decides', so the LLM cannot execute an effect the kernel has not authorized. Runs locally on your own Claude or Hermes key, with no paid service dependency.
Claude knowledge-work-plugins· tool, 2026-09-17 Apache-2.0 repo from Anthropic: 11 open-source, forkable starter plugins that turn Claude into a role specialist, covering sales, marketing, finance, legal, support, product, data, enterprise search, bio-research and productivity. Each is a manifest plus MCP tool connectors, slash commands, and skills Claude draws on. You install them into Claude Cowork or the Claude Code CLI and customize with your own context; they run inside Claude, not as standalone apps.
Rowboat· tool, 2026-09-17 Apache-2.0 TypeScript app from Rowboat Labs: a self-hostable multiplayer workspace for teams where each person brings their own AI assistant, and those assistants collaborate across shared Spaces with local knowledge. The Spaces server (Harbor) is open source and self-hostable; the desktop app runs against local models via Ollama or LM Studio, or hosted models with your own key. Optional voice and search integrations (Deepgram, ElevenLabs, Exa) take their own keys.
BrowserSkill· tool, 2026-09-16 MIT-licensed Rust tool from Tencent that connects an AI agent to your real, already-logged-in Chrome instead of a fresh headless one: a local CLI plus a browser extension let the agent borrow one of your open tabs, drive it with your existing cookies and sessions, and then hand control back to you. The agent works inside the auth state you already have, so it can reach logged-in pages without you scripting a login or pasting credentials into a bot.
Bumblebee· tool, 2026-09-16 Apache-2.0 Go tool from Perplexity that inventories the packages and editor extensions installed on a developer machine and checks them against catalogs of known-compromised software, so a supply-chain compromise gets caught on the workstation rather than in production. It is a single static binary with zero non-stdlib dependencies, installed with `go install` and run locally; the compromised-software catalogs ship with it.
Open Code Review· tool, 2026-09-16 Apache-2.0 AI code-review CLI that Alibaba ran internally for ~two years before open-sourcing it: it reads your git diff, runs the changed files through a hybrid deterministic-pipeline + LLM-agent pass with a built-in multi-language ruleset (null-pointer, thread-safety, XSS, SQLi), and posts line-level comments. Bring your own OpenAI- or Anthropic-compatible key, or use delegation mode so a coding agent's own model does the review. Installs from npm.
OpenWiki· tool, 2026-09-16 MIT-licensed TypeScript CLI from LangChain that writes and maintains agent-readable documentation of your codebase, so a coding agent has persistent context instead of re-deriving your repo on every task. It keeps state locally in ~/.openwiki and works against local repositories; it can run through local model backends (Ollama, LM Studio, any OpenAI-compatible endpoint) or hosted providers, so you are not forced onto a paid key.
agentdesktop· tool, 2026-09-15 Open-source visibility and control layer for AI tools across a desktop fleet (Apache-2.0, Rust), backed by Solo.io: it discovers every harness, model, MCP server and skill on a machine (per device and user), applies central policy that reconciles drift, and treats agents as non-human identities issued short-lived credentials scoped to device, user and tool. Runs standalone from local YAML, or in fleet mode via an MDM daemon wired to your identity provider and LLM gateway.
Vibe· tool, 2026-09-15 Open-source, fully-offline transcription desktop app (MIT, Rust) built on OpenAI's Whisper models: transcribe audio and video locally in 90+ languages, detect speakers, translate, and capture your microphone or system audio, with no account and no data leaving the machine. Exports SRT, VTT, TXT, DOCX, PDF and JSON for subtitle and content workflows. Runs on macOS, Windows and Linux, GPU-accelerated when available.
AfterShip/email-verifier· tool, 2026-09-15 MIT-licensed Go library that verifies email addresses without sending mail: syntax checks, MX records, SMTP lookups with catch-all detection, auto-updated disposable-domain lists, role-account detection, typo suggestions and SOCKS5 proxy support. Positions itself as broader-featured than competing Go libraries.
Atomic CRM· tool, 2026-09-15 MIT-licensed full-featured CRM built with React, shadcn-admin-kit and Supabase: contact management, kanban deal pipeline, tasks with reminders, email capture via CC and multi-provider auth (Google, Azure, Keycloak, Auth0). Its fork-first model lets you customize deeply yet pull upstream component updates via the shadcn registry. An open alternative to Attio.
Frappe Builder· tool, 2026-09-15 MIT-licensed low-code visual website builder for designers and developers: craft and instantly publish sites in the Frappe ecosystem. Its AI site builder, Bob, grounds itself in your site — reading existing pages and reusing your components and theme tokens rather than guessing; frappe.io itself runs on it. An open alternative to Webflow.
jBPM/Apache KIE· tool, 2026-09-15 Apache-2.0 umbrella (Apache Incubator) for integrated business automation: the Drools rule engine (DMN, complex event processing), jBPM workflow engine (BPMN 2.0), OptaPlanner constraint solver and cloud-native Kogito runtime. The oldest open-source BPM lineage — relevant as Camunda 7 reaches EoL and Camunda 8's core went source-available.
Puck· tool, 2026-09-15 MIT-licensed modular visual editor for React: build custom drag-and-drop page-building experiences from your own React components. It ships as just a React component — works in any React environment including Next.js, with no vendor lock-in and your data staying yours (~13.3k stars).
Temporal· tool, 2026-09-15 Durable-execution platform (MIT core) for distributed applications: the runtime persists workflow state, handles intermittent failures and retries failed operations automatically, so business logic survives process crashes. Originated as a fork of Uber's Cadence by its creators; ships a Web UI, CLI and multi-language SDKs. Industry standard for the pattern — but heavy to operate self-hosted.
Wexflow· tool, 2026-09-15 MIT-licensed, cross-platform workflow automation engine (C#): automate file operations, system processes, scripting, networking, email and scheduled batch runs with 100+ built-in activities, a web dashboard, Docker support and 6+ database engines. Approval workflows are first-class; a single-maintainer project.
DeepScrape· tool, 2026-09-14 Open-source web scraper (MIT, TypeScript) that turns pages into agent-readable data: Playwright automation plus a fit-markdown extractor (pruning content filters) for clean Markdown, and an LLM-extraction path (GPT-4o) that returns structured JSON to a schema. Ships a hardened one-command Docker deployment (managed-Redis ready, non-root). A smaller, self-hostable entry in the URL-to-Markdown category alongside Firecrawl, Crawl4AI and browser-use.
Voicebox· tool, 2026-09-14 Open-source, fully-local AI voice studio (MIT, Python + React) from Jamie Pine: clone a voice from a short sample, generate speech across 7 TTS engines and 23 languages, and dictate into any text field with a global hotkey (a Wispr Flow alternative). The GTM hook is one MCP tool call, voicebox.speak, that lets any MCP-aware agent talk back in a voice you cloned. Models, voice data and captures never leave the machine; runs on Apple Silicon (MLX) or CUDA/ROCm/CPU (PyTorch), macOS and Windows.
DeepSeek Harness (dsh)· tool, 2026-09-13 Open-source agent harness from DeepSeek AI (MIT, TypeScript) where everything is a plugin: the model adapter, tool registry, sandbox and agent loop are all swappable, so it runs any model rather than one vendor's. Ships four agent presets, an instant web UI (npx @deepseek-ai/dsh web), OS-level sandboxing (Landlock/Seatbelt/ACL), full MCP support, and a subagent bridge that plugs Claude Code and Codex in as tools. Currently v0.1 (RC7), a developer preview.
Anomalib· tool, 2026-09-13 Apache-2.0 deep-learning library from the Open Edge Platform (Intel lineage) for visual anomaly detection in images and video: Lightning-based, ready-to-use model implementations from experimentation through deployment, with OpenVINO export for accelerated inference. Claims the largest public collection of anomaly-detection algorithms; the bundled Anomalib Studio web app is pre-release.
Billion Mail· tool, 2026-09-13 Fully self-hosted, open-source (AGPL-3.0) email server and email-marketing platform: newsletters, campaigns and transactional mail on your own server, free of monthly per-subscriber fees. Vendor claims an 8-minute install to first sent email. An open alternative to Mailchimp (~15.6k stars).
Coneshare· tool, 2026-09-13 MIT-licensed, self-hosted alternative to DocSend and virtual datarooms: an access-control, engagement-analytics and workflow layer over files that never leave your existing storage (Nextcloud, Google Drive, Dropbox) — down to page-by-page reading time and video playback. Early-stage, and effectively the only open-source entrant in digital-sales-room reading.
enthec/webappanalyzer· tool, 2026-09-13 GPL-3.0 database of regex technology fingerprints — HTML, JS variables, headers, cookies, DOM, DNS and meta tags — for detecting what a website runs. The community continuation of Wappalyzer after it went private in August 2023, with an explicit commitment to stay open source. A technographics backbone for prospecting and security tooling.
Frappe CRM· tool, 2026-09-13 Open-source (AGPL-3.0) CRM for modern sales teams — leads, deals and customer interactions, positioned against CRMs it calls too complex with steep pricing. Built-in telephony: make, receive and record calls inside the CRM via Twilio and Exotel. Self-hostable; an open alternative to Attio.
toot· tool, 2026-09-13 ~1323 stars; complements Mastodon.py; OAuth2 login against open Mastodon API; CLI + TUI. Open protocol — no ban/ToS risk. GPL-3.0 copyleft (matters if embedding/redistributing, not for CLI use).
zg (zvec-grep)· tool, 2026-09-13 A local-first search tool for people and AI agents that combines ripgrep, BM25, and on-device vector search across code, docs, and data — indexing and search stay on the machine. Auto-configures agents (Claude Code, Codex, OpenCode) and reports fewer tool-calls/tokens on SWE-QA-Bench.
OpenExecutive· tool, 2026-09-13 A self-hostable virtual executive advisory board: one executive voice backed by eight specialist agents (strategy, finance, HR, legal, ops, marketing, product, board comms), with RAG over your own company docs, episodic memory across sessions, and a follow-up scheduler. Ships Slack/email/Telegram/Discord/web/CLI.
AI Engineering Coach· tool, 2026-09-12 Open-source VS Code extension from Microsoft (MIT; also runs in Cursor and Antigravity) that reads your LOCAL AI-coding session logs — GitHub Copilot, Claude Code, Codex CLI, OpenCode, Xcode — and scores your workflow across prompt quality, session hygiene, code review, tool mastery and context management, with 45 anti-pattern detection rules. Read-only, zero telemetry.
litelm· tool, 2026-09-12 Open-source Python library (MIT, 182★) that extracts the core of LiteLLM — model routing, message translation, streaming, tool use and embeddings across 19 providers — into ~2,900 lines with two dependencies (openai, httpx), keeping API compatibility with LiteLLM for drop-in migration.
LLM Wiki· tool, 2026-09-12 Open-source desktop app (GPL-3.0, 19.1k★; Tauri, TypeScript + Rust) that compiles your documents into a persistent, interconnected wiki instead of re-reading them per query like standard RAG: two-step chain-of-thought ingest, a knowledge graph with community detection, cited multi-conversation chat, web clipper, and Obsidian-vault compatibility.
offer-builder· tool, 2026-09-12 Open-source Claude Code skill (MIT) that turns dense market material — industry reports, transcripts, research — into a cold-email-ready offer: it identifies the real problem and its cost, then produces an Offer Sheet with market analysis, value proposition, elevator pitch and a targeting matrix for list building. Deliberately stops before copywriting.
OpenVurp· tool, 2026-09-12 Open-source, local-first multi-agent framework (MIT, Python): create named agents with specific jobs that run on your computer, use real tools, read files, search the web and consult each other, reachable over Telegram, Discord, Slack and WhatsApp — with sandboxing and approval workflows built in, and local models via Ollama or LM Studio.
Reef· tool, 2026-09-12 Open-source infrastructure (Apache-2.0, 1.1k★ within ~10 days of release) for continually self-improving agents: a serve → observe → grow → commit loop that matches production feedback to updates and ships them as versioned changes — either training model weights (via Slime/SGLang) or evolving the agent's harness of prompts, rules and skills.
VoiceStudio· tool, 2026-09-12 Open-source, fully-local voice AI suite (AGPL-3.0, 23.8k★): voice cloning from one reference clip, video dubbing across 646 languages, audiobooks, dictation and transcription, wrapping 14 TTS engines. No account, API key or usage meter for the local workflow — the ElevenLabs alternative that keeps audio on your machine.
Cline (Desktop App)· tool, 2026-09-11 Open-source coding agent (Apache-2.0, 67.8k★) that just shipped a standalone macOS/Windows desktop app positioned for open-weight models: the established bring-your-own-model agent now runs outside the IDE, pairing naturally with local or hosted open-weight LLMs instead of a single vendor's frontier API.
DearAgent· tool, 2026-09-11 Open-source email infrastructure for AI agents (AGPL-3.0, TypeScript, self-hosted on your Cloudflare account): agents create their own addresses, receive mail, wait for verification codes, reply in threads and send — exposed via REST API, webhooks and an MCP server, so an agent can run a real inbox instead of borrowing yours.
Hyper-τ-Bench· tool, 2026-09-11 Open-source benchmark (MIT, Python) from Sierra Research that tests whether an AI coding agent can BUILD a working customer-service agent — not just operate one: the developer agent gets a sandbox with business records, a codebase and a production API, and must design, wire and ship a functioning agent scored against simulated customers.
Raggy· tool, 2026-09-11 Open-source CLI (MIT, Python) for retrieval-augmented Q&A over your local documents: hybrid retrieval (vector + BM25), cross-encoder reranking, and fully local embedding/indexing via Ollama — answer generation can stay local or use a remote API, so private GTM docs never have to leave your machine.
Rune· tool, 2026-09-11 Rune, a fast GPU-rendered keyboard-driven IDE (Go) that folds code, terminals, CLI tools, language intelligence, debugging and AI agents into one multi-workspace environment, just went open source under GPL-3.0 — a polished commercial-grade dev environment moving to a copyleft license.
Portal AI Plugins (Spotify shunt)· tool, 2026-09-11 Open-source Claude Code plugin set (Apache-2.0) from Spotify's Portal platform: the 'shunt' plugin routes I/O-heavy agent work — bulk file reads and boilerplate generation — to cheaper worker models via PreToolUse hooks, cutting Claude token usage by ~90% in Spotify's benchmarks on a Java monorepo.
Relay· tool, 2026-09-10 Open-source push-to-talk voice interface layer (React/TypeScript) for dense operations dashboards: deterministic intent parsing, a set of UI components and a reusable skill so an AI coding agent can add voice control over analytics views. VERY EARLY.
Security Cards· tool, 2026-09-10 Open-source security-guidance dataset and installable agent skill (Astro site + data): targeted, version-aware security best practices for 80+ widely-used libraries across 13 languages, aimed at coding agents that write functionally-correct-but-insecure code. In the makers' BaxBench eval, wiring it into Claude Code (Opus 4.7) cut the insecure-code rate from 23.9% to 6.6%.
Stroq· tool, 2026-09-10 Open-source local action firewall (TypeScript) for AI coding agents: it intercepts an agent's operations, tracks session state by tainting what the agent has read, and blocks dangerous follow-up actions before they run on your machine — a runtime guardrail with intent/audit that complements install-time scanners.
Cynative· tool, 2026-09-09 Open-source framework (Go) for building your own security agents: ships 45 built-in agents that examine AWS, GCP, Azure, Kubernetes, GitHub and GitLab for privilege escalation, public exposure and supply-chain risks using frontier models. Cloud-security-focused — adjacent to, not core, GTM.
Emailclaw· tool, 2026-09-09 Open-source email AI agent (Java) that uses the email subject line as a project boundary to plan and run multi-step tasks with strict per-workspace isolation — an inbox-native automation surface. VERY EARLY: single-digit stars.
Geiger· tool, 2026-09-09 Open-source, dependency-free CLI (JavaScript) that inventories every AI agent, MCP server, plugin and extension on a machine and reports what each can access — read-only, no install, no telemetry. Agent observability and least-privilege review for a local stack.
MagicVault· tool, 2026-09-09 Open-source secrets layer (Rust) that lets agents use enrolled credentials without ever seeing them: it injects secrets into approved browser fields, HTTP requests or spawned processes while keeping them out of prompts, replies and audit logs. VERY EARLY.
Mnemiq· tool, 2026-09-09 Open-source text-to-SQL engine (Python) that turns natural-language questions into queries against your own database, with a configurable pipeline, built-in safety checks and integrated evaluation tooling — the analytics/BI node for a self-hosted agent stack.
OpenPost· tool, 2026-09-09 Open-source, self-hostable social-media management platform (Go core; SvelteKit web, Expo mobile): create, adapt, schedule and publish content across networks with built-in image/video editing and analytics — an AGPL-3.0 alternative to Buffer, CapCut and Canva that replaces three GTM SaaS bills with one self-hosted stack.
TeamAI CLI· tool, 2026-09-09 Open-source CLI (TypeScript) from Tencent that keeps a whole team's AI-agent configuration in sync from one shared, version-controlled repo: skills, rules, MCP servers and knowledge propagate automatically to every teammate's Claude Code, Cursor, Codex, OpenCode and other agents — no manual copy-paste, not locked to one agent.
Tripwire· tool, 2026-09-09 Open-source, sandboxed security scanner (Python) for AI skills and MCP servers: discovers and scans the exact skill files and MCP servers you are about to install and surfaces findings in one dashboard, with pluggable scanners. The supply-chain-review layer for an agent stack.
Headroom· tool, 2026-09-08 Open-source context-compression proxy for AI agents (Rust core; Python/TypeScript SDKs): sits between your agent and the LLM and compresses tool outputs, logs, JSON, code and RAG chunks with format-specific, reversible compressors before the model sees them — no changes to agent code.
Substack MCP· tool, 2026-09-08 Open-source MCP server (TypeScript) for Substack: search your archive, create and review drafts, publish Notes, read analytics and manage subscribers across publications — exposing content operations to an AI agent as tools plus a CLI (the maker cites 24 tools).
Crew· tool, 2026-09-07 Open-source coordination layer that lets parallel coding-agent sessions (Claude Code, Codex, opencode) share live context, exchange messages and split work without branching.
aiogram· tool, 2026-09-03 ~5800 stars; complements python-telegram-bot, Telethon; Official Bot API framework (FSM, middleware, magic filters). Auth via bot token. No ToS risk. Complementary alt to python-telegram-bot.
airtable.js· tool, 2026-09-03 ~2223 stars; complements pyairtable; VENDOR (Airtable) official JavaScript client for the Airtable REST API (PAT/OAuth). ToS-clean and authoritative for Node/browser.
ALwrity· tool, 2026-09-03 Full-stack content OS w/ GSC SEO; solo-maintained
Apache SeaTunnel· tool, 2026-09-03 OSS alt to Fivetran/Airbyte; ~9543 stars; via ossinsight.io
Apache Unomi· tool, 2026-09-03 Genuine OSI CDP; ASF-governed; heavy Java
AppFlowy· tool, 2026-09-03 OSS alt to Notion; ~75102 stars; AGPL copyleft; via ossinsight.io
Astro· tool, 2026-09-03 GATE PASS; modern content framework
Async PRAW· tool, 2026-09-03 ~154 stars; complements PRAW, redditwarp; aiohttp-based async twin of PRAW; same OAuth2 sanctioned Reddit API and rate-limit compliance. ToS-clean official approach for asyncio workloads.
atproto (python-sdk)· tool, 2026-09-03 ~654 stars; complements skyware-js/bot, mattn/bsky; App-password / OAuth session against the open AT Protocol (XRPC), auto-generated from official lexicons; sync+async. Open protocol — no ToS/ban risk; PDS-agnostic. De-facto Python SDK for Bluesky.
Bento· tool, 2026-09-03 OSS alt to closed stream processors; the fully-open Benthos; ~2100 stars; complements Benthos/redpanda-data-connect (upstream), Trigger.dev, SeaTunnel (XREF); MIT. Community fork of Benthos, created after the original (now redpanda-data/connect) moved to a restrictive source-available license -> Bento is the OSI-open continuation. Config-driven stream pipelines using official service creds. No per-platform ToS flag.
Bifrost· tool, 2026-09-03 OSS alt to LiteLLM; ~7151 stars; via ossinsight.io
bolt-python (Slack Bolt)· tool, 2026-09-03 ~1300 stars; complements python-slack-sdk; Official Slack Bolt framework (Events API, Socket Mode, slash commands). Auth via bot/app tokens + OAuth. No ToS risk.
BrightBean Studio· tool, 2026-09-03 Unified inbox+sentiment; influencer discovery has NO OSS
bsky (mattn)· tool, 2026-09-03 ~454 stars; complements atproto (python-sdk), skyware-js/bot; Go CLI over the open AT Protocol using app-password auth. Substituted for the ambiguous 'bsky-cli' name (several exist); this is the most-adopted active CLI. Open protocol — no ban/ToS risk.
BusinessDirectory· tool, 2026-09-03 Modern Next.js; no monetization/claim; citation-mgmt has NO OSS
caldav (python-caldav)· tool, 2026-09-03 ~411 stars; complements vdirsyncer, khal; Open-protocol CalDAV client (works with Nextcloud, iCloud, Google via CalDAV). Auth via basic/app-password/OAuth per server. Open standard — no ToS/ban risk. Dual GPL-3.0/Apache-2.0.
CAMEL· tool, 2026-09-03 OSS alt to CrewAI; ~17562 stars; via ossinsight.io
cdp-use· tool, 2026-09-03 Pure Chrome DevTools Protocol, type-safe in Python; browser-automation primitive; ~308 stars
CLI-Anything· tool, 2026-09-03 Framework and registry for making existing software agent-native through structured CLIs; seven-phase harness generator, CLI-Hub package manager, JSON output, tests, SKILL.md generation, and maintained harnesses for creative, productivity, engineering, GIS, media, and developer applications; ~47.6k stars; generated CLIs execute the permissions of their wrapped software — review before agent use
CocoIndex· tool, 2026-09-03 OSS alt to Pipedream; ~11205 stars; via openalternative.co
Cognee· tool, 2026-09-03 OSS alt to LlamaIndex; ~29874 stars; via ossinsight.io
Composio· tool, 2026-09-03 OSS alt to Zapier-for-agents, proprietary tool platforms; ~29600 stars; complements ACI.dev, Nango, Pipedream, MCP servers; Root LICENSE is MIT (some managed-server subdirectories carry Elastic-2.0). Handles OAuth/managed-auth to 250+ apps and exposes them as agent tools/MCP; uses each app's official credentials -> ethical_risk empty. Cloud offering is separate.
cua (Computer-Use Agents)· tool, 2026-09-03 OSS alt to proprietary computer-use sandboxes; ~21100 stars; complements macOS-use (XREF), self-operating-computer, Open Interpreter; MIT. NOT the same as XREF macOS-use: this is the broader trycua/cua project (Computer/Lume/Lumier + cross-OS agent fleets & benchmarks) with sandboxed VMs. Sandboxing is the intended safety boundary for computer-use. No per-platform ToS flag.
discord.py· tool, 2026-09-03 ~16100 stars; complements serenity; Official Discord bot API wrapper. Auth via bot token. No ToS risk for bots (self-botting user accounts IS banned - don't). Proper rate limiting built in.; via ossinsight.io, GitHub cross-ref
discourse_api· tool, 2026-09-03 ~304 stars; complements pydiscourse (Python client); Official Discourse (discourse org) Ruby client for the sanctioned API. Auth via admin API key. No ToS/ban risk.
Dittofeed· tool, 2026-09-03 OSS alt to Customer.io; ~2888 stars; via openalternative.co
dlt· tool, 2026-09-03 Pure OSI Python EL library; schema inference
DVC· tool, 2026-09-03 OSS alt to proprietary MLOps asset stores; ~15800 stars; complements git-lfs, ComfyUI workflow files; Closest fit for versioning large generated assets/models alongside Git. Branch is otherwise thin — no purpose-built gen-asset versioning product exists. No model.
ERPNext· tool, 2026-09-03 OSS alt to SAP/NetSuite; ~37793 stars; broad ERP; CRM is one module; via ossinsight.io
Fediwall· tool, 2026-09-03 Only UGC-curation dashboard; Fediverse-only
gcalcli· tool, 2026-09-03 ~3800 stars; complements google-api-python-client (official-api); CLI over the sanctioned Google Calendar API. Auth via Google OAuth (own client credentials). No ToS risk.
gh (GitHub CLI)· tool, 2026-09-03 ~45800 stars; complements PyGithub, octokit.js; Official GitHub CLI. Auth via OAuth device flow or PAT. No ToS risk; standard REST rate limits (5000/hr authenticated).
glab (GitLab CLI)· tool, 2026-09-03 ~4200 stars; complements python-gitlab; Official GitLab CLI (now under gitlab-org). Auth via PAT/OAuth. No ToS risk. Repo lives on GitLab, not GitHub.
gmailctl· tool, 2026-09-03 ~2200 stars; complements google-api-python-client; Uses official Gmail API to sync filters. Auth via OAuth2. Diff-before-apply. No ToS risk.
GoAccess· tool, 2026-09-03 OSS alt to Google Analytics; ~20763 stars; via ossinsight.io
google-api-python-client (Docs/Drive)· tool, 2026-09-03 ~8900 stars; complements gmailctl; Official Google API client (maintenance mode but supported). Gmail API auth via OAuth2. No ToS risk; per-user API quotas.
Google Workspace CLI· tool, 2026-09-03 Official one-command CLI for Drive/Gmail/Calendar/Docs; Rust; ~30k stars; scriptable Workspace automation
googlemaps (google-maps-services-python)· tool, 2026-09-03 ~5000 stars; complements google-maps-scraper (unofficial fallback); Official Google Maps Platform Python client (Geocoding, Places, Directions, Distance Matrix). Auth via API key. Sanctioned; no ToS risk; billed per request.
Gorse· tool, 2026-09-03 Only ready-to-deploy engine; REST API + dashboard
Gradio· tool, 2026-09-03 ~43314 stars; ML-focused; BI fit is secondary; via ossinsight.io
GrapesJS· tool, 2026-09-03 GATE PASS; framework not finished product
gspread· tool, 2026-09-03 ~7492 stars; complements google-api-python-client; Ergonomic wrapper over the official Google Sheets API (OAuth2 / service account). ToS-clean; the go-to Python Sheets client. Subject to Sheets API per-minute read/write quotas.; via GitHub cross-ref, ossinsight.io
Hermes Agent· tool, 2026-09-03 Nous Research self-improving agent; persistent memory + autonomous skill creation; multi-LLM-provider; ~228k stars (anomalously high — verify hype); CAVEAT: autonomous skill-creation = arbitrary code execution by design — sandbox before use
Higress· tool, 2026-09-03 ~9048 stars; via ossinsight.io
himalaya· tool, 2026-09-03 ~7000 stars; complements imap_tools, isync/mbsync; Open protocols (IMAP/SMTP/JMAP) plus Gmail REST and Graph backends. Auth: password/OAuth2/XOAUTH2. No ToS risk.
imap_tools· tool, 2026-09-03 ~836 stars; complements himalaya; Open IMAP protocol (over imaplib). Auth via credentials/app-password/OAuth. No ToS risk.
Inspect AI· tool, 2026-09-03 UK AISI; rigorous eval (verified)
jira-cli· tool, 2026-09-03 ~5900 stars; complements pycontribs/jira, go-jira; Uses official Jira REST API. Auth via API token (Cloud) or PAT (Server/DC). No ToS risk.
jira (pycontribs)· tool, 2026-09-03 ~2100 stars; complements jira-cli, go-jira; De-facto Python Jira library (Cloud + Server/DC). Auth via API token/basic/OAuth. Official API; no ToS risk.
JupyterLab· tool, 2026-09-03 Standard notebook host; use v4
KeystoneJS· tool, 2026-09-03 OSS alt to Contentful; ~9938 stars; via openalternative.co
khal· tool, 2026-09-03 ~3000 stars; complements caldav, vdirsyncer; Standards-based terminal calendar; syncs to CalDAV servers via vdirsyncer. Open protocol — no ToS/ban risk.
Lowcoder· tool, 2026-09-03 Retool-clone successor; NocoDB reclassified SA
mail-parser (SpamScope)· tool, 2026-09-03 ~452 stars; complements email-reply-parser; Pure parsing library (no network/auth). RFC-compliant, Outlook .msg support, defect detection. No ToS risk.
Mastodon.py· tool, 2026-09-03 ~957 stars; complements toot; OAuth2 app tokens against the documented, open Mastodon REST API (open protocol / ActivityPub). No ToS/ban risk — self-hostable, per-instance rate limits. Feature-complete through Mastodon 4.5.
matplotlib· tool, 2026-09-03 Custom license — flag for strict scanners
matrix-nio· tool, 2026-09-03 ~705 stars; complements mautrix-python; Open Matrix client-server protocol with E2EE. Auth via access token/password. Open federated standard; no ToS risk.
mautrix-python· tool, 2026-09-03 ~251 stars; complements matrix-nio; Matrix framework powering the mautrix bridges (and maubot). Open protocol; appservice auth via registration token. No ToS risk on Matrix side (individual bridges to WhatsApp/Telegram carry those platforms' ToS).
MCP Inspector· tool, 2026-09-03 ~10621 stars; via ossinsight.io
MCPJungle· tool, 2026-09-03 OSS alt to proprietary MCP gateways; ~1200 stars; complements MetaMCP, lasso mcp-gateway; MPL-2.0 (weak copyleft). Self-hosted registry+gateway: register servers once, expose one endpoint to Claude/Cursor/Codex/agents with access control. Auth delegated to registered servers. No per-platform ToS flag.
MetaMCP· tool, 2026-09-03 OSS alt to proprietary MCP gateways; MCP Context Forge (XREF); ~2600 stars; complements MCPJungle, lasso mcp-gateway, Klavis; MIT (successor to metatool-app). Groups MCP servers into namespaces, exposes unified SSE/HTTP/OpenAPI endpoints with middlewares + API-key/OIDC auth, Docker-packaged. It fronts other servers; auth is delegated to them. No per-platform ToS flag.
modelcontextprotocol/servers· tool, 2026-09-03 OSS alt to proprietary agent-tool plugins; ~89000 stars; complements Klavis, awesome-mcp-servers, MCP Context Forge (XREF); Reference servers under MIT, transitioning to Apache-2.0 (new contributions Apache-2.0; unrelicensed MIT contributions stay MIT) -> OSI-open either way. Each server uses the target service's official auth. No per-platform ToS flag.
msgraph-sdk-python (Microsoft Graph)· tool, 2026-09-03 ~624 stars; complements python-o365, exchangelib; Official Microsoft Graph v1.0 Python SDK. Auth via MSAL OAuth2 (app/delegated). No ToS risk.
notesmd-cli (Obsidian CLI Community)· tool, 2026-09-03 ~1568 stars; complements obsidian-git; Go CLI acting on LOCAL Obsidian vault markdown files (plus obsidian:// URIs) — no platform account, no ToS/ban risk (local files). Active; formerly named obsidian-cli. Finalist for local-KB automation.
notion-sdk-js· tool, 2026-09-03 ~5649 stars; complements notion-sdk-py, ultimate-notion; VENDOR (makenotion) official JavaScript/TypeScript SDK for the Notion API (integration token / OAuth). ToS-clean and authoritative for Node projects.
notion-sdk-py· tool, 2026-09-03 ~2174 stars; complements notion-sdk-js, ultimate-notion; Sync+async Python client mirroring the OFFICIAL Notion API (integration token / OAuth). ToS-clean; the de-facto Python SDK. Standard Notion API rate limit ~3 req/s per integration.; via GitHub cross-ref, ossinsight.io
Observable Plot· tool, 2026-09-03 Concise grammar-of-graphics on D3
octokit.js· tool, 2026-09-03 ~7800 stars; complements gh CLI, PyGithub; Official GitHub JS SDK (Node/Deno/Browser). Auth via token/OAuth/App. No ToS risk.
Ollama· tool, 2026-09-03 Easiest local runtime; don't expose port publicly
Open Interpreter· tool, 2026-09-03 OSS alt to ChatGPT Code Interpreter (local); ~67900 stars; complements cua, self-operating-computer, PyAutoGUI; Apache-2.0. LLM writes+runs code locally to operate the machine (incl. computer-use / OS control). Executes arbitrary code -> sandbox or require confirmation; not a per-platform ToS risk but a local-security risk.
OpenAlternative· tool, 2026-09-03 Only deployable alt-finder app; +CC0 data list
OpenCheck· tool, 2026-09-03 KYB tool over registry DATA; no contact emails
OpenFlow· tool, 2026-09-03 Only lead-capture-specific OSS; young; rest collapse into covered form/page builders
osmnx· tool, 2026-09-03 ~5800 stars; complements python-overpy (raw Overpass), yelpapi; Open-data OSM tooling over the public Overpass/Nominatim APIs. Open protocol/open data (ODbL) — no ToS/ban risk; respect Overpass usage policy.
Payload CMS· tool, 2026-09-03 OSS alt to Contentful/Strapi; ~44044 stars; via ossinsight.io
pgvector· tool, 2026-09-03 Cleanest profile; no telemetry/CVEs; Postgres extension
Plotly.py· tool, 2026-09-03 Interactive notebook viz; ties to Dash
pyairtable· tool, 2026-09-03 ~892 stars; complements airtable.js; Python client for the OFFICIAL Airtable REST API (personal access token / OAuth). ToS-clean; de-facto Python SDK. Airtable API rate limit ~5 req/s per base.
Pydantic AI· tool, 2026-09-03 OSS alt to LangGraph; ~19148 stars; via ossinsight.io
PyGithub· tool, 2026-09-03 ~7800 stars; complements gh CLI, octokit.js; Official-API client. Auth via PAT/OAuth/GitHub App. LGPL-3.0 copyleft (weak). No ToS risk; REST rate limits apply.
PyOD· tool, 2026-09-03 ~9953 stars; via ossinsight.io
python-amazon-sp-api· tool, 2026-09-03 ~671 stars; complements amzn/selling-partner-api-models (official OpenAPI models); De-facto Python client for the sanctioned Amazon SP-API. Auth via LWA OAuth + AWS IAM/role. No ToS risk (official API); respect SP-API usage plans/rate limits.
python-asana· tool, 2026-09-03 ~335 stars; complements py-trello, clickupython, monday-sdk-js; Official Asana SDK (300+ endpoints). Auth via PAT/OAuth. No ToS risk.
python-gitlab· tool, 2026-09-03 ~2500 stars; complements glab CLI; Official-API Python client, bundles a CLI. Auth via PAT/OAuth. LGPL-3.0. No ToS risk.
python-o365· tool, 2026-09-03 ~1900 stars; complements msgraph-sdk-python, exchangelib; Ergonomic wrapper over Microsoft Graph. Auth via OAuth2. No ToS risk; friendlier than raw Graph SDK.
python-slack-sdk· tool, 2026-09-03 ~4000 stars; complements bolt-python; Official Slack SDK (Web API, Socket Mode, SCIM, audit logs). Auth via tokens/OAuth. No ToS risk. Lower-level than Bolt.
python-telegram-bot· tool, 2026-09-03 ~29400 stars; complements Telethon, aiogram; De-facto standard on official Telegram Bot API. Auth via bot token. No ToS risk (sanctioned Bot API). LGPL-3.0.; via ossinsight.io, GitHub cross-ref
pythreads· tool, 2026-09-03 ~72 stars; complements threads-py, threads-api; OAuth2 against Meta's SANCTIONED Threads API (Graph-style) — covers all published endpoints. ToS-clean official approach; the correct modern replacement for the dead reverse-engineered libs. Requires a Meta developer app.
quizco· tool, 2026-09-03 Only real OSS quiz builder; education-positioned not lead-gen
Ray· tool, 2026-09-03 OSS alt to proprietary GPU-orchestration platforms; ~43500 stars; complements Celery, SkyPilot; General distributed-compute engine; also serves the cost/GPU/model orchestration branch. No model.
Recommenders· tool, 2026-09-03 LF-governed best-practices toolkit; not turnkey service
Refferq· tool, 2026-09-03 Most permissive; payout automation
Relaticle· tool, 2026-09-03 OSS alt to Attio; ~1494 stars; via openalternative.co
Robot Framework· tool, 2026-09-03 OSS alt to UiPath Studio, Blue Prism; ~11800 stars; complements rpaframework (Robocorp), TagUI, Selenium/Playwright; Apache-2.0. Mature de-facto OSS RPA/test framework; deterministic keyword-driven (not LLM). No auth model of its own; inherits credentials of the libraries it drives. No per-platform ToS flag.
rpaframework (Robocorp)· tool, 2026-09-03 OSS alt to UiPath, Automation Anywhere activity packs; ~1500 stars; complements Robot Framework, TagUI; Apache-2.0. Battery of RPA libraries (Browser, Excel, PDF, Email, Windows, cloud) for Robot Framework + plain Python. Deterministic. Credentials handled by each library; no per-platform ToS flag.
ScrapeGraphAI· tool, 2026-09-03 OSS alt to Diffbot; ~29238 stars; via opensource.builders
Scrapy· tool, 2026-09-03 Mature; Zyte-backed; no built-in JS render
Sendune· tool, 2026-09-03 OSS alt to Resend; ~2182 stars; via openalternative.co
SEOnaut· tool, 2026-09-03 Web-app dashboard; forkable to SaaS
serenity· tool, 2026-09-03 ~5600 stars; complements discord.py; Official Discord API in Rust (sharding, cache, gateway). Auth via bot token. No ToS risk for bots.
Shopify CLI· tool, 2026-09-03 ~729 stars; complements shopify_python_api (official-api); Official Shopify CLI. OAuth login via Partners account. Sanctioned tooling; no ToS risk.
shopify_python_api· tool, 2026-09-03 ~1400 stars; complements Shopify CLI (cli); Official Shopify SDK. Auth via API key / OAuth access token. Sanctioned API, no ToS risk; respect Admin API rate limits (leaky-bucket).
signal-cli· tool, 2026-09-03 ~4800 stars; complements signald; Unofficial CLI built on official libsignal-client; registers/links a real Signal account. Widely relied-upon de-facto standard for server notifications -> finalist despite unofficial. platform-tos flagged (unsanctioned client; number could be limited). Auth via phone registration/link.
Silex· tool, 2026-09-03 GATE PASS-with-copyleft-caveat; clean HTML export
SiteOne Crawler· tool, 2026-09-03 Rust; CI gates; backlink slice has NO OSS (data moat)
SkyPilot· tool, 2026-09-03 OSS alt to proprietary GPU-cost platforms; ~10500 stars; complements Ray; Provisions cheapest GPUs across clouds with spot/failover — the strongest OSS fit for the cost/GPU branch (STRUCTURAL: most cost-optimization is commercial). No model.
social-post· tool, 2026-09-03 Multi-platform; hard dep on MuAPI vendor
spaCy· tool, 2026-09-03 Production NLP; Explosion-maintained
StackStorm· tool, 2026-09-03 Rules+workflow engine; heavy infra
Stanza· tool, 2026-09-03 SOTA neural accuracy; 60+ languages
stripe-cli· tool, 2026-09-03 ~2100 stars; complements stripe-python (official-api); Official Stripe CLI. Auth via API key/OAuth device login. Read/config + webhook testing; never automate money movement without human approval.
stripe-python· tool, 2026-09-03 ~2000 stars; complements stripe-cli (cli); Official Stripe SDK. Auth via secret API key. READ/CONFIG ONLY for this atlas: never automate money movement (charges/refunds/transfers/payouts) without explicit human approval. Use restricted keys and test mode.
STUMPY· tool, 2026-09-03 ~4140 stars; via ossinsight.io
Tiledesk· tool, 2026-09-03 Live chat + LLM bot builder + human handoff
Timeplus Proton· tool, 2026-09-03 OSS alt to Materialize; ~2238 stars; via openalternative.co
trackthenews· tool, 2026-09-03 Media monitoring; PR distribution/journalist-DB has NO OSS
tweepy· tool, 2026-09-03 ~11178 stars; complements twikit, twscrape, snscrape; OAuth 1.0a / OAuth 2.0 + API key/bearer token against sanctioned X API v2. ToS-compliant; subject to paid API tiers and per-tier rate limits. No ban risk when used with an authorized developer app.; via GitHub cross-ref, ossinsight.io
workflow-use· tool, 2026-09-03 'RPA 2.0' — record + replay deterministic browser workflows on top of browser-use; ~4k stars; AGPL network-copyleft (SaaS caution)
WritingTools· tool, 2026-09-03 Local rewrite/tone; no detector-evasion framing
X Algorithm· tool, 2026-09-03 Official source for the X For You recommendation stack: candidate retrieval, engagement prediction, diversity-aware ranking, content/account labeling, visibility filtering, ads blending, and supporting Rust/JAX services; ~31.4k stars; valuable reference implementation but not a standalone deployable social platform and some anti-abuse rules are intentionally withheld
Model Context Protocol Registry· tool, 2026-09-03 Community-driven registry service for Model Context Protocol servers. Provides a stateless REST read API (v0.1) with cursor pagination and updated_since delta sync, plus the server.json entry schema that standardizes tool distribution metadata.
Firecrawl anydoc· tool, 2026-09-03 Open-source document-to-Markdown converter from the Firecrawl team (Rust with multi-language bindings): Word, PowerPoint, Excel, OpenDocument, RTF, EPUB, CSV and PDF into clean GitHub-Flavoured Markdown — an ingest layer for agent and RAG pipelines.
OpenBot· tool, 2026-09-02 Gives each AI agent its own isolated computer (browser, file workspace, tools) behind a policy gateway that reviews and audit-logs every action before it runs. Works with AG-UI agents (LangGraph, Mastra, CrewAI, Pydantic AI); model-agnostic and self-hostable.
UI-TARS Desktop / Agent TARS· tool, 2026-09-02 An open multimodal GUI-automation agent stack that controls a computer via a vision-language model — clicking, typing, and reading the screen locally — with a CLI and Web UI, browser control, and MCP-tool integration. From ByteDance.
FreeToken· tool, 2026-09-01 Edge-native Mixture-of-Experts inference engine (UC Berkeley) that runs frontier open-weight models on consumer hardware — e.g. a 35B model on an 8GB GPU — by treating GPU/CPU/host-memory as one bandwidth-adaptive platform; 2–4× faster than Ollama on MoE. Desktop app + Python CLI.
SIE (Superlinked Inference Engine)· tool, 2026-09-01 Self-hosted, OpenAI-compatible inference server that runs 85–100+ open models behind one API (encode/score/extract/generate), loading and LRU-evicting models by traffic so one GPU serves a rotating set instead of one server per model. Plugs into Qdrant, Weaviate, Chroma, LanceDB, LangChain, LlamaIndex.
Obscura· tool, 2026-08-31 Open-source headless browser for AI agents and web scraping, built in Rust: runs V8 JavaScript, speaks Chrome DevTools Protocol, drop-in for Puppeteer/Playwright, ~30MB, with stealth/anti-fingerprinting and MCP support — no Chromium.
URL Intelligence Agent· tool, 2026-08-31 An evidence-first enrichment agent that turns a public URL into structured, sourced intelligence: entity resolution, contact/team extraction, tech-stack detection, brand assets, SEO/security signals, and relationship graphs — with confidence scores and provenance. Runs as CLI, HTTP API, MCP server, or self-hosted service.
daily-briefing· tool, 2026-08-29 Open-source AI news-briefing agent that runs entirely on GitHub Actions with no server: Groq for the model + Tavily for web search + email delivery. A reference pattern for owned, near-zero-cost GTM automation.
Itsuki· tool, 2026-08-29 Open-source memory engine for AI agents over REST API + MCP: captures conversations/events into typed, versioned, reversible objects (facts, entities, relationships, events) accessible across 26+ tools, with an audit trail and JSON export.
Tokensift· tool, 2026-08-29 Open-source, tokenizer-level static linter for LLM prompts and payloads — flags token waste (UUIDs, repeated blocks, pretty-printed JSON, base64 blobs) with fix suggestions. Library + CLI, 20 built-in rules; exact counts for OpenAI, calibrated estimates for Claude.
advertools· tool, 2026-08-25 Python productivity toolkit for online marketing: SEO crawling, SERP analysis, keyword/SEM tools, robots.txt & sitemap parsing, log-file and search-marketing analysis.
Chrome DevTools MCP· tool, 2026-08-25 MCP server that lets an agent inspect and control a live Chrome browser — performance traces, network requests, console errors, rendered pages, and screenshots.
Docling· tool, 2026-08-25 Document-processing toolkit with advanced PDF/DOCX/PPTX/XLSX understanding (tables, complex layouts) and generative-AI integrations — for when simple PDF-to-text loses structure.
Google Ads API Python Client· tool, 2026-08-25 Google's official Python client for the Google Ads API — build reporting, budget monitoring, anomaly detection, and controlled account-management (write-capable; gate mutations behind approval).
Google Analytics MCP Server· tool, 2026-08-25 The official Google Analytics team's MCP server: run GA4 core/funnel/realtime reports and property queries through the Analytics read scope — one of the safer direct marketing-data connections.
Google Agent Skills· tool, 2026-08-25 Google's official collection of Agent Skills across infrastructure, databases, AI/ML, security, and developer tools (incl. Google Ads/Analytics developer workflows). Installable via `npx skills add google/skills`. Broad, not marketing-only.
GPT Researcher· tool, 2026-08-25 Autonomous research agent that produces cited, multi-source reports from web and local documents — useful for market, competitor, and category research and content briefs. MCP integration included.
Marketing Skills· tool, 2026-08-25 A collection of AI-agent skills for marketing tasks — CRO, copywriting, SEO, paid ads, analytics, launches, pricing, and lifecycle/growth marketing.
MarkItDown· tool, 2026-08-25 Microsoft utility that converts PDFs, Word, PowerPoint, Excel, HTML, images and more into clean Markdown for LLM pipelines — structure-preserving.
Meridian· tool, 2026-08-25 Google's media-mix-modeling (MMM) framework for measuring channel contribution and modeling budget allocation — privacy-safe marketing science, not ad-copy generation.
Playwright MCP· tool, 2026-08-25 Microsoft MCP server that lets agents automate a browser via structured accessibility snapshots (not screenshots) — for rendered-page checks, forms, funnels, and landing-page QA.
PyMC-Marketing· tool, 2026-08-25 Bayesian marketing toolbox in PyMC: media-mix modeling (MMM), customer lifetime value (CLV), and buy-till-you-die (BTYD) retention models.
Robyn (MMM)· tool, 2026-08-25 Meta's experimental marketing-mix-modeling package — models diminishing returns, response curves, and budget allocation across channels using ML.
TikTok Business API SDK· tool, 2026-08-25 TikTok's official SDK (Java/Python/JS) for its Business API — auth, campaign creation, reporting, and Business Center workflows. Apply the same test-account + approval rules to mutations.
Milvus for DeepSeek Harness· tool, 2026-08-25 A read-only DeepSeek Harness plugin that lets an agent inspect and query Milvus / Zilliz Cloud vector stores — entity lookup, scalar queries, BM25 full-text, and dense+BM25 hybrid retrieval. The plugin bounds the tool surface (no insert/delete/admin) at the execution boundary, not by prompt.
oc (only-cli)· tool, 2026-08-25 Open-source CLI that fetches a web page and returns a compact numbered view instead of raw HTML — a cheap web-read layer for scraping/enrichment agents. Installs as a CLI and as an agent skill (npx skills add).
Quartermaster (QM)· tool, 2026-08-25 An open, multiplayer agent harness for organizations: every employee and Slack room gets a scoped workspace (separate memory, files, permissions, sandbox), with shared skills, admin-gated promotion, crons/background watches, and one-command deploy to Fly/AWS. Model-agnostic (Claude/Codex/OpenCode).
SkillOpt· tool, 2026-07-02 SkillOpt is a Microsoft Research optimizer that trains reusable natural-language skills for frozen LLM agents, keeping only trajectory-driven edits that pass benchmark validation and transferring the result across models. Open source: yes (MIT); self-hostable; pricing free. Python, PyPI, 17.6k stars.