dox
Open-source agentic AI framework that runs autonomous agents on your own machine with Docker sandboxing and any LLM provider.
Pick Agent Zero if you want autonomous agents with real filesystem and terminal reach, a readable log of every action, and the freedom to swap in any LLM, including local models via Ollama or LM Studio. The v2.9 security hardening, the A0 CLI Connector's Docker-plus-host-terminal design, and the announced v3 gVisor sandbox direction show the project treats sandboxing as a first-class concern rather than an afterthought. Skip it if you want a hosted, click-to-run assistant — the self-hosting, Docker-based setup is the entry fee — or if the bundled A0T token governance layer is a dealbreaker you cannot ignore.
Verified 8d ago · liveness 64/100 · cite: rightaichoice.com/tools/dox
- Developers building and customizing autonomous agents they fully own
- Security researchers running automated penetration testing and vulnerability analysis
- Data analysts building MCP-powered pipelines from web and Salesforce sources
- Researchers who need custom knowledge management and agentic RAG
- Non-technical users who want a no-code, click-to-run AI assistant
- Teams that need a fully managed cloud service with no self-hosting
- Beginners without command-line or Docker experience
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Agent Zero if you want a hosted, click-to-run assistant and are not comfortable running Docker on your own machine — the self-hosting and configuration work is the entry fee, not a temporary obstacle.
You pay your own model provider's API rates: Agent Zero itself is open source and free to run, but every agent step that calls a cloud LLM bills to your OpenRouter, OpenAI, Anthropic, or other API account.
Agent Zero is open source and free to run; your real recurring cost is the model provider you connect it to, which for local Ollama or LM Studio setups can be near zero. That puts it well under managed agent platforms like Manus or per-seat assistant subscriptions such as ChatGPT, Claude, and Gemini — but it asks for the Docker and configuration work those hosted tools absorb for you.
In short
dox — Open-source agentic AI framework that runs autonomous agents on your own machine with Docker sandboxing and any LLM provider. Best for Developers building and customizing autonomous agents they fully own, Security researchers running automated penetration testing and vulnerability analysis, Data analysts building MCP-powered pipelines from web and Salesforce sources. Free to use.
What's new in dox
Checked 8 days agoAcross the latest 4 updates: 3 feature updates and 1 news mention.
Agent Zero v2.11: Context Doctor, Vision Sidecars, & Faster Streaming
v2.11 adds Context Doctor for smarter context management, vision sidecars for multimodal processing, and faster streaming, plus WebUI synchronization improvements.
Agent Zero v2.10: Interactive Browser, Durable Subordinates, & Composer Context
v2.10 introduces an interactive shared browser runtime, durable subordinate lifecycles, scoped composer context and drafts, and a built-in ACP bridge.
Agent Zero v3: A Sneak Peek Inside the Sandbox
Agent Zero's third major iteration is moving agent-run code and heavyweight tools into isolated, disposable gVisor sandboxes.
Agent Zero v2.9: Agent Editor, Scoped Policies, & Security Hardening
v2.9 introduces the Agent Editor, project-scoped profiles and tool policies, safer profile lifecycles, Time Travel retention, and built-in plugin upgrades.
What people actually say about dox — is it worth it?
We scanned public community sources for dox on Aug 30, 2026 and could not establish that the discussion we found is about this tool rather than something else sharing its name. Our own analysis of that scan says the posts were off-subject. Rather than publish a sentiment score built on the wrong subject, we publish nothing here and re-run the scan.
Viability Score
How well maintained and how widely used is dox? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Autonomous agents with local terminal, file, and browser access
- Dynamic tool creation and reuse across runs
- Docker sandbox execution with host access through A0 CLI Connector
- Step-by-step workflow logging for full audit trails
- A0 Launcher desktop app for install, updates, backup, and multi-instance management
- Plugin Hub with one-click install and AI-driven security scans
- Connect any AI provider via API including Ollama and LM Studio
- Free private AI API key via Venice AI included
- Agentic RAG for memory and knowledge management
- Context Doctor for context management (v2.11)
- Vision sidecars for multimodal processing (v2.11)
- Interactive shared browser runtime (v2.10)
- Durable subordinate agent and subagent lifecycles (v2.10)
- Agent Editor with project-scoped profiles and tool policies (v2.9)
- Memory Dashboard for storing, retrieving, and managing agent memories
About dox
Agent Zero is an open-source agentic AI framework that runs autonomous agents on your own computer or inside a Docker sandbox, with zero secrets exposed to a closed platform. It targets developers, security researchers, data analysts, researchers, and content creators who want agents with real terminal, file, and browser access. The core loop is unusual: the agent gets a computer, a full toolset, and the ability to write new tools when it needs them. It learns from past runs, self-corrects, and logs every step for audit. The A0 Launcher desktop app handles installation, updates, backup of /a0/usr, and multi-instance management from one place, while the A0 CLI Connector keeps agents sandboxed in Docker yet able to reach your host terminal, local files, and browser. You can connect practically any AI provider through its API, including local models via Ollama or LM Studio, and a free private API key through Venice AI is included. Recent releases added Context Doctor for tighter context management and vision sidecars for multimodal work (v2.11), an interactive shared browser runtime and durable subordinate lifecycles (v2.10), and the Agent Editor with project-scoped profiles and tool policies (v2.9). v2.13 redesigned file navigation, added project folders and messaging slash commands. Governance runs on the A0T token on Ethereum Base L2. Compared with managed agent platforms like Manus or general assistants like ChatGPT, Claude, and Gemini, Agent Zero trades convenience for ownership: you host it, you inspect it, you extend it.
Behind the Verdict
Agent Zero's differentiator is the environment, not the model. Most agent tools are a chat box with function calling bolted on; Agent Zero gives the agent an actual computer and lets it write new tools when the built-in set runs out — then reuses them on later runs. Everything is logged step by step, which is the feature that matters most if you are running automated penetration tests or data pipelines and need to explain what happened afterward. The A0 CLI Connector is the clever architectural piece: the agent stays sandboxed in Docker while reaching your real terminal, project files, and browser, so you get host-level capability without giving up isolation. The A0 Launcher handles install, updates, backup of /a0/usr, and multi-instance management, which removes the worst rough edges of self-hosting a Python-based framework. Strengths stack up around ownership and transparency — you control the provider, the prompts under usr/agents, the tool policies per project, and the memory via the Memory Dashboard and agentic RAG layer. The Plugin Hub adds one-click community extensions with AI-driven security scans before deployment, and you can package your own skills, tools, and extensions into a plugin and publish it. On the weakness side, this is not a no-code product. You need command line and Docker comfort, and you need to think about the vendor's recommended Launcher, first-run onboarding with OpenRouter as default, model presets, and project-scoped secrets. The bundled crypto governance layer — A0T on Ethereum Base L2 with staking, topic voting, and Improvement Proposals — is genuine functionality, but if you do not care about token governance it is noise you can ignore rather than remove. Where it fits: security teams wanting automated sweeps they can audit, developers building reusable custom agents, researchers who need custom knowledge management and agentic RAG over their own sources, and anyone running agents on data that cannot leave local infrastructure. Where it does not: teams that want a fully managed cloud service, beginners without Docker experience, and buyers expecting a large library of prebuilt enterprise SaaS connectors out of the box.
Researching dox? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas dox actually fits — and what changes day-one when you adopt it.
Install Agent Zero through the A0 Launcher, connect a provider via the first-run onboarding wizard, then point an agent at your lab infrastructure for an automated sweep. The agent writes new tools when its built-in set runs short, and every action is written to the step-by-step log.
Outcome: You get a repeatable penetration-test run with an auditable trace you can hand to a client, and the same agent profile can be rerun next quarter against the same project.
Create a Project with its own instructions, files, memory, and secrets, then connect Salesforce and web sources over MCP so the agent can extract and reconcile data on a schedule set in the Task Scheduler.
Outcome: Pipelines run on your own machine on a schedule, source credentials stay in project-scoped secrets, and past runs feed the agentic RAG memory so the agent improves on repeated jobs.
Wrap the custom skills, tools, and extensions you built into a plugin, run it through the Plugin Hub's AI-driven security scan, and install it in another Agent Zero instance with one click.
Outcome: Your internal agents become portable plugins your teammates can install from the UI instead of copying code, and the security scan gives the team a reviewable check before deployment.
Use Cases
- Run automated penetration tests and vulnerability sweeps on your own infrastructure with a dedicated agent that logs every step.
- Build MCP-powered data pipelines that extract and analyze data from web and Salesforce sources.
- Create and edit technical documentation or blog posts with AI assistance under project-scoped files and memory.
- Pull real-world stock data from Yahoo Finance and turn raw prices into actionable insights.
- Deploy an autonomous research agent that conducts literature reviews and gathers insights into a persistent memory store.
- Develop and package custom Skills, tools, and plugins you can publish to the Plugin Index.
- Run recurring scheduled tasks as subagents for long-running jobs that survive across sessions.
Limitations
- Agent Zero runs on your own computer and connects to AI providers via API, including local models through Ollama or LM Studio, with a free private API key via Venice AI.
- Installation goes through either the guided Launcher desktop app or the CLI, and managing instances and backing up /a0/usr involves real technical steps.
- The project's own documentation is the primary reference, and the tool's value assumes you can run Docker.
- The bundled A0T token governance layer on Ethereum Base L2 adds staking, topic voting, and Improvement Proposals that many buyers will not need.
- Features like the interactive browser runtime require host-browser mode and per-project configuration, so setup is not a one-click affair.
as of 2026-09-30
Verification history
We have re-verified dox 9 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 9 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Where the pricing makes sense
The company stage and team size where dox's pricing actually pencils out — and where peers do it cheaper.
Agent Zero is open source and free to run; your real recurring cost is the model provider you connect it to, which for local Ollama or LM Studio setups can be near zero. That puts it well under managed agent platforms like Manus or per-seat assistant subscriptions such as ChatGPT, Claude, and Gemini — but it asks for the Docker and configuration work those hosted tools absorb for you.
Setup time & first value
How long it actually takes to get something useful out of dox — broken out by persona, not the marketing-page minute.
Developers comfortable with Docker get to a running first agent in roughly 30 to 60 minutes: install the A0 Launcher, let it set up Docker, complete first-run onboarding with OpenRouter or a local provider, and open the Web UI. Security researchers and analysts should budget a few hours to configure projects, secrets, tool policies, and the plugins or MCP connections a specific workflow needs
Switching to or from dox
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From a managed agent platform like Manus: recreate your workflow as an Agent Zero project with its own files, memory, and secrets, then connect your preferred provider through the first-run onboarding wizard.
- →From a local script-based agent setup: package your scripts as Agent Zero tools or Skills so the agent can call them dynamically instead of you invoking them by hand.
- →From ChatGPT, Claude, or Gemini: move your reusable instructions into an Agent Profile under usr/agents and point the agent at local files and a browser via the A0 CLI Connector.
- ↗To a hosted agent platform: export your project files and agent memories from the Web UI and rebuild the workflow against the host's own connector library.
- ↗To a plain scripting setup: package your Agent Zero Skills and custom tools as standalone scripts, since the tools live as editable files under your instance directory.
Integrations
Resources & Guides
Tutorials & Learning
YouTube returned 6 videos for “dox”, and we withheld 6: 6 could not be judged, because “dox” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about dox.
Official links
Tools that pair well with dox
Common stack mates teams adopt alongside dox, with the specific reason each pairing earns its keep.
AutoGen
Microsoft's open-source framework for building conversational and event-driven AI agents in Python and .NET.
Mastra
Open-source TypeScript framework for building durable AI agents and workflows, with a hosted platform for observability and cloud deployment.
Zhipu GLM
Zhipu GLM (Z.ai) ships the open-weights GLM-5.3 family, full-modality MaaS APIs, and autonomous agents like AutoGLM and GLM-PC.
Featured Head-to-Head Comparisons
Dox vs Presto Voice
Presto Voice and dox serve completely different needs: Presto is a specialized drive-thru voice AI for QSR chains aiming to boost revenue and efficiency, while dox is an open-source agentic AI framework for developers who want full control over autonomous computer tasks. Your choice depends entirely on whether you run a restaurant chain (Presto) or build custom AI agents (dox).
Dox vs Truleo
If you run a law enforcement agency drowning in siloed systems and manual case research, Truleo is the specialized AI intelligence platform that delivers automated leads and cuts report writing from 40 minutes to 7 minutes. But if you're a developer wanting to build and control autonomous AI agents on your own computer for free, dox (Agent Zero) offers unmatched flexibility and transparency. There's no overlap—choose based on your sector and technical comfort.
Dox vs Locus Robotics
Locus Robotics and dox serve entirely different markets: Locus is a physical AMR solution for warehouse automation (RaaS), while dox is an open-source AI agent framework for autonomous computer tasks. Choose Locus if you need to scale physical order fulfillment in high-volume warehouses; choose dox if you are a developer wanting to run autonomous agents on your own machine with full control. There is no overlap in use cases.
Alternatives to dox
View all
Frequently Asked Questions
Used dox? Help shape our editorial sentiment research.