Updated Oct 23, 9:23 AM Bogotá
Agent harnesses & personal agents
Open agent harnesses / personal AI agents you can run yourself.
live · Refreshed every 24 hours · Updated Oct 23, 9:23 AM Bogotá · 49% of cells are filled (633 of 1242 have no value yet).
Compare: Harnesses that run without Docker · Harnesses with Telegram · Harnesses with scheduled tasks and helper agents · Harnesses that shorten long conversations for you
Hidden by default: 61 more column, including Website, Repository, Main language. Use “All columns” or the chooser to see them.
3 entries have no checked fact yet, so they are not shown here until they do.
Choose…
Name is always shown: the chooser lists the other columns.
How this data is built · What each column means · Column names and values are in English: the directory keeps one data standard, so a value means the same thing in every language version.
Where these values come from
Per column: how many values, from which sites, when they were last checked, and one full link you can open. Every value in the table links to its own source.
- License: 15 values from github.com (newest check Oct 22, 11:58 AM Bogotá) · 15 found automatically, none checked by a person yet. https://github.com/openclaw/openclaw
- Main language: 16 values from github.com (newest check Oct 22, 11:58 AM Bogotá) · 16 found automatically, none checked by a person yet. https://github.com/openclaw/openclaw
- GitHub stars: 18 values from github.com (newest check Oct 22, 1:44 PM Bogotá) · 18 found automatically, none checked by a person yet. https://github.com/openclaw/openclaw
- Last code change: 18 values from github.com (newest check Oct 22, 11:58 AM Bogotá) · 18 found automatically, none checked by a person yet. https://github.com/openclaw/openclaw
- Latest release: 17 values from github.com (newest check Oct 22, 1:44 PM Bogotá) · 17 found automatically, none checked by a person yet. https://github.com/openclaw/openclaw/releases/tag/v2026.10.6
- Latest release date: 17 values from github.com (newest check Oct 22, 1:44 PM Bogotá) · 17 found automatically, none checked by a person yet. https://github.com/openclaw/openclaw/releases/tag/v2026.10.6
- Needs Docker: 9 values from docs.paperclip.ing, picoclaw.io, raw.githubusercontent.com and 1 more sites (newest check Oct 22, 2:36 PM Bogotá) · 9 found automatically, none checked by a person yet. https://raw.githubusercontent.com/openclaw/openclaw/main/README.md
- Where it runs: 18 values from bloks.dev, craftbot.live, docs.nanoclaw.dev and 5 more sites (newest check Oct 22, 2:40 PM Bogotá) · 7 checked by a person, 11 found automatically. https://github.com/openclaw/openclaw
- Pricing model: 17 values from bloks.dev, codync.dev, github.com and 2 more sites (newest check Oct 22, 2:51 PM Bogotá) · 10 checked by a person, 7 found automatically. https://github.com/openclaw/openclaw
- Software price ($): 17 values from appubot.com, bloks.dev, codync.dev and 3 more sites (newest check Oct 22, 2:51 PM Bogotá) · 11 checked by a person, 6 found automatically. https://github.com/openclaw/openclaw
- Software price per month ($/month) (calculated): 17 calculated from other columns.
- Hosted plan offered: 4 values from appubot.com, bloks.dev, craftbot.live and 1 more sites (newest check Oct 22, 2:51 PM Bogotá) · 4 checked by a person, 0 found automatically. https://docs.letta.com/pricing
- Hosted plan from ($/month): 3 values from bloks.dev, craftbot.live, docs.letta.com (newest check Oct 22, 2:51 PM Bogotá) · 3 checked by a person, 0 found automatically. https://docs.letta.com/pricing
- Hosted plan up to ($/month): 2 values from bloks.dev, craftbot.live (newest check Oct 22, 2:51 PM Bogotá) · 2 checked by a person, 0 found automatically. https://craftbot.live/pricing
- Free plan: 7 values from appubot.com, bloks.dev, craftbot.live and 4 more sites (newest check Oct 22, 2:51 PM Bogotá) · 5 checked by a person, 2 found automatically. https://raw.githubusercontent.com/openclaw/openclaw/main/README.md
- AI usage fees: 4 values from appubot.com, bloks.dev, docs.letta.com and 1 more sites (newest check Oct 22, 2:51 PM Bogotá) · 4 checked by a person, 0 found automatically. https://opencode.ai/zen
- Vendor AI plan from ($/month): 2 values from bloks.dev, opencode.ai (newest check Oct 22, 2:51 PM Bogotá) · 2 checked by a person, 0 found automatically. https://opencode.ai/go
- Vendor AI plan up to ($/month): 2 values from bloks.dev, opencode.ai (newest check Oct 22, 2:51 PM Bogotá) · 2 checked by a person, 0 found automatically. https://opencode.ai/go
- Flat fee ($): 2 values from docs.letta.com, opencode.ai (newest check Oct 22, 3:42 PM Bogotá) · 2 checked by a person, 0 found automatically. https://opencode.ai/zen
- Flat fee charged per: 2 values from docs.letta.com, opencode.ai (newest check Oct 22, 3:42 PM Bogotá) · 2 checked by a person, 0 found automatically. https://opencode.ai/zen
- Which AI models: 14 values from docs.paperclip.ing, github.com, murage.ai and 2 more sites (newest check Oct 22, 2:40 PM Bogotá) · 6 checked by a person, 8 found automatically. https://github.com/openclaw/openclaw
- Built-in AI providers: 13 values from docs.paperclip.ing, github.com, murage.ai and 2 more sites (newest check Oct 22, 2:41 PM Bogotá) · 5 checked by a person, 8 found automatically. https://github.com/openclaw/openclaw
- Number of built-in AI providers (calculated): 13 calculated from other columns.
- Runs local models: 10 values from github.com, murage.ai, opencode.ai and 1 more sites (newest check Oct 22, 2:38 PM Bogotá) · 5 checked by a person, 5 found automatically. https://github.com/openclaw/openclaw
- Any OpenAI-compatible service: 6 values from github.com, murage.ai, raw.githubusercontent.com (newest check Oct 22, 1:50 PM Bogotá) · 2 checked by a person, 4 found automatically. https://github.com/NousResearch/hermes-agent
- Any Anthropic-compatible service: 1 values from pi.dev (newest check Oct 22, 2:34 PM Bogotá) · 1 found automatically, none checked by a person yet. https://pi.dev/docs/latest/models-configuration
- Uses coding-agent tools: 8 values from github.com, murage.ai, raw.githubusercontent.com (newest check Oct 22, 2:40 PM Bogotá) · 2 checked by a person, 6 found automatically. https://github.com/openclaw/openclaw
- AI models included: 5 values from appubot.com, bloks.dev, craftbot.live and 2 more sites (newest check Oct 22, 2:51 PM Bogotá) · 5 checked by a person, 0 found automatically. https://opencode.ai/zen
- Use your own keys: 16 values from bloks.dev, docs.paperclip.ing, github.com and 2 more sites (newest check Oct 22, 1:49 PM Bogotá) · 6 checked by a person, 10 found automatically. https://github.com/openclaw/openclaw
- Chat apps: 18 values from docs.letta.com, docs.picoclaw.io, hermes-agent.nousresearch.com and 5 more sites (newest check Oct 22, 2:39 PM Bogotá) · 18 found automatically, none checked by a person yet. https://raw.githubusercontent.com/openclaw/openclaw/main/README.md
- Phone chat apps (calculated): 17 calculated from other columns.
- Desktop app: 11 values from hermes-agent.nousresearch.com, openclaw.ai, raw.githubusercontent.com (newest check Oct 22, 1:37 PM Bogotá) · 11 found automatically, none checked by a person yet. https://openclaw.ai
- Conversation style: 12 values from docs.letta.com, docs.openclaw.ai, lorca.app and 3 more sites (newest check Oct 22, 8:59 AM Bogotá) · 1 checked by a person, 11 found automatically. https://docs.openclaw.ai/concepts/session
- Shortens long chats itself: 4 values from docs.openclaw.ai, opencode.ai, raw.githubusercontent.com (newest check Oct 22, 8:59 AM Bogotá) · 1 checked by a person, 3 found automatically. https://docs.openclaw.ai/concepts/session
- Memory stored as: 6 values from docs.letta.com, docs.openclaw.ai, github.com and 2 more sites (newest check Oct 22, 2:36 PM Bogotá) · 4 checked by a person, 2 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory shared by: 9 values from appubot.com, docs.letta.com, docs.openclaw.ai and 3 more sites (newest check Oct 22, 2:40 PM Bogotá) · 4 checked by a person, 5 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory writer: 8 values from appubot.com, docs.letta.com, docs.openclaw.ai and 3 more sites (newest check Oct 22, 2:40 PM Bogotá) · 4 checked by a person, 4 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory cleanup: 2 values from docs.openclaw.ai, github.com (newest check Oct 22, 2:29 PM Bogotá) · 2 checked by a person, 0 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory cleanup every (hours): 3 values from docs.openclaw.ai, github.com, raw.githubusercontent.com (newest check Oct 22, 2:36 PM Bogotá) · 2 checked by a person, 1 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory recall: 6 values from docs.openclaw.ai, github.com, hermes-agent.nousresearch.com and 1 more sites (newest check Oct 22, 2:38 PM Bogotá) · 4 checked by a person, 2 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory size limit (tokens): 1 values from hermes-agent.nousresearch.com (newest check Oct 22, 2:29 PM Bogotá) · 1 checked by a person, 0 found automatically. https://hermes-agent.nousresearch.com/docs/user-guide/features/memory
- Memory kept when chat is shortened: 3 values from docs.openclaw.ai, hermes-agent.nousresearch.com, raw.githubusercontent.com (newest check Oct 22, 2:39 PM Bogotá) · 2 checked by a person, 1 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory kept after a fresh start: 6 values from docs.letta.com, docs.openclaw.ai, github.com and 2 more sites (newest check Oct 22, 2:39 PM Bogotá) · 4 checked by a person, 2 found automatically. https://docs.openclaw.ai/concepts/memory
- Keeps a profile of you: 7 values from docs.openclaw.ai, github.com, hermes-agent.nousresearch.com and 1 more sites (newest check Oct 22, 2:40 PM Bogotá) · 3 checked by a person, 4 found automatically. https://docs.openclaw.ai/concepts/memory
- Keeps a project log: 1 values from docs.openclaw.ai (newest check Oct 22, 2:29 PM Bogotá) · 1 checked by a person, 0 found automatically. https://docs.openclaw.ai/concepts/memory
- Saves skills: 10 values from docs.letta.com, docs.openclaw.ai, github.com and 2 more sites (newest check Oct 22, 2:40 PM Bogotá) · 3 checked by a person, 7 found automatically. https://docs.openclaw.ai/tools/skills
- Memory readable by you: 6 values from docs.letta.com, docs.openclaw.ai, hermes-agent.nousresearch.com and 1 more sites (newest check Oct 22, 2:40 PM Bogotá) · 2 checked by a person, 4 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory editable by you: 7 values from docs.letta.com, docs.openclaw.ai, github.com and 2 more sites (newest check Oct 22, 2:40 PM Bogotá) · 3 checked by a person, 4 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory exportable: 3 values from docs.openclaw.ai, hermes-agent.nousresearch.com, raw.githubusercontent.com (newest check Oct 22, 2:36 PM Bogotá) · 2 checked by a person, 1 found automatically. https://docs.openclaw.ai/concepts/memory
- Memory history kept: 4 values from docs.letta.com, raw.githubusercontent.com (newest check Oct 22, 2:39 PM Bogotá) · 4 found automatically, none checked by a person yet. https://docs.letta.com/concepts/shared-memory
- Writes its own memory (calculated): 8 calculated from other columns.
- Learns about you: 7 values from docs.letta.com, docs.openclaw.ai, raw.githubusercontent.com (newest check Oct 22, 2:33 PM Bogotá) · 1 checked by a person, 6 found automatically. https://docs.openclaw.ai/concepts/memory
- Writes its own skills: 6 values from docs.letta.com, docs.openclaw.ai, raw.githubusercontent.com (newest check Oct 22, 9:02 AM Bogotá) · 1 checked by a person, 5 found automatically. https://docs.openclaw.ai/tools/skills
- Scheduled tasks: 12 values from craftbot.live, docs.letta.com, picoclaw.io and 1 more sites (newest check Oct 22, 2:37 PM Bogotá) · 12 found automatically, none checked by a person yet. https://raw.githubusercontent.com/paperclipai/paperclip/master/README.md
- Schedules its own tasks: 5 values from craftbot.live, docs.openclaw.ai, raw.githubusercontent.com (newest check Oct 22, 8:59 AM Bogotá) · 1 checked by a person, 4 found automatically. https://docs.openclaw.ai/automation/cron-jobs
- Helper agents: 12 values from appubot.com, craftbot.live, docs.letta.com and 2 more sites (newest check Oct 22, 1:36 PM Bogotá) · 12 found automatically, none checked by a person yet. https://raw.githubusercontent.com/anomalyco/opencode/dev/README.md
- Agent teams: 10 values from appubot.com, lorca.app, nanobot.wiki and 1 more sites (newest check Oct 22, 1:37 PM Bogotá) · 10 found automatically, none checked by a person yet. https://raw.githubusercontent.com/paperclipai/paperclip/master/README.md
- Messages you first: 9 values from appubot.com, docs.openclaw.ai, lorca.app and 2 more sites (newest check Oct 22, 2:35 PM Bogotá) · 1 checked by a person, 8 found automatically. https://docs.openclaw.ai/automation/cron-jobs
- MCP tools: 14 values from docs.letta.com, opencode.ai, pi.dev and 1 more sites (newest check Oct 22, 1:50 PM Bogotá) · 14 found automatically, none checked by a person yet. https://raw.githubusercontent.com/NousResearch/hermes-agent/main/README.md
- Memory use when idle (MB): 1 values from github.com (newest check Oct 22, 2:29 PM Bogotá) · 1 checked by a person, 0 found automatically. https://github.com/sipeed/picoclaw
- Summary: 18 values from lorca.app, opencode.ai, raw.githubusercontent.com (newest check Oct 22, 2:40 PM Bogotá) · 18 found automatically, none checked by a person yet. https://raw.githubusercontent.com/openclaw/openclaw/main/README.md
- Tags: 18 values from docs.letta.com, opencode.ai, raw.githubusercontent.com (newest check Oct 22, 2:40 PM Bogotá) · 18 found automatically, none checked by a person yet. https://raw.githubusercontent.com/openclaw/openclaw/main/README.md
What each column means
- Name: The project's name. Source: entered by hand.
- Website: The project's official website. Source: entered by hand.
- Repository: The project's GitHub repository, written as owner/name. Source: entered by hand.
- License: The open-source license as reported by GitHub, for example MIT or Apache-2.0. Source: GitHub.
- Main language: The main programming language of the code, as reported by GitHub. Source: GitHub.
- GitHub stars: How many people have starred the repository on GitHub. Source: GitHub.
- Last code change: Date of the most recent commit to the repository. Source: GitHub.
- Latest release: Version name of the newest release published on GitHub. Source: GitHub.
- Latest release date: Date the newest GitHub release was published. Source: GitHub.
- Needs Docker: Yes when the documented install requires Docker. Source: the project's own pages, read by AI and checked against our rules.
- Where it runs: Whether you run it on your own machine, the vendor runs it for you, or both are possible. Possible values: Your machine, Vendor's cloud, Both. Source: the project's own pages, read by AI and checked against our rules.
- Pricing model: How the software itself is paid for, such as free open source or a subscription. Possible values: Free open source, Subscription, Pay per use, One-time lifetime license, Percent of volume, Network cost plus flat fee. Source: the project's own pages, read by AI and checked against our rules.
- Software price ($): What the software itself costs and how often you pay it; $0 for free open source. Unit: US dollars. Source: the project's own pages, read by AI and checked against our rules.
- Software price per month ($/month) (calculated): The software price in US dollars per month, calculated from the Software price column: monthly prices as is, yearly prices divided by 12, $0 stays $0, and other one-time prices are left empty. Source: calculated from other columns.
- Hosted plan offered: Yes when the vendor offers to run it for you as a hosted service. Source: the project's own pages, read by AI and checked against our rules.
- Hosted plan from ($/month): Monthly price of the cheapest hosted plan, in US dollars. Unit: US dollars. Source: the project's own pages, read by AI and checked against our rules.
- Hosted plan up to ($/month): Monthly price of the most expensive listed hosted plan, in US dollars. Unit: US dollars. Source: the project's own pages, read by AI and checked against our rules.
- Free plan: Yes when there is a $0 plan or a free usage allowance. Source: the project's own pages, read by AI and checked against our rules.
- AI usage fees: How AI model usage is charged: not at all, passed through at cost, or with a markup. Possible values: Not at all, At cost (no markup), With markup. Source: the project's own pages, read by AI and checked against our rules.
- AI usage markup (%): The percentage the vendor adds on top of the model provider's price. Unit: percent. Source: the project's own pages, read by AI and checked against our rules.
- Vendor AI plan from ($/month): Monthly price of the cheapest subscription to AI models sold by the vendor itself, such as OpenCode Go. Unit: US dollars. Source: the project's own pages, read by AI and checked against our rules.
- Vendor AI plan up to ($/month): Monthly price of the most expensive subscription to AI models sold by the vendor itself. Unit: US dollars. Source: the project's own pages, read by AI and checked against our rules.
- Flat fee ($): A fixed fee in US dollars, charged once per the unit shown in the Flat fee charged per column. Unit: US dollars. Source: the project's own pages, read by AI and checked against our rules.
- Flat fee charged per: What the flat fee is charged on, such as each $20 credit top-up or each active agent per month. Possible values: each payment, each $20 credit top-up, each active agent per month, per call, per seat per month. Source: the project's own pages, read by AI and checked against our rules.
- Which AI models: Which AI models you can use, one of: the vendor's own models only; many models through the vendor's account; any OpenAI-compatible service; your keys, only for providers the app already lists. Possible values: Vendor's models only, Many models via vendor account, Any OpenAI-compatible service, Your keys, only for providers the app already lists. Source: the project's own pages, read by AI and checked against our rules.
- Built-in AI providers: AI model providers it supports out of the box. Possible values: OpenAI, Anthropic, Google, xAI, OpenRouter, Ollama, OpenCode, Groq, Together AI, DeepSeek, Mistral, llama.cpp (local), LM Studio, vLLM, Azure OpenAI, AWS Bedrock, Google Vertex, Nous Portal, Zhipu (GLM), Moonshot (Kimi), Qwen, MiniMax, GitHub Copilot, Fireworks, Cerebras, Perplexity, Hugging Face, Venice. Source: the project's own pages, read by AI and checked against our rules.
- Number of built-in AI providers (calculated): How many providers are listed, counted from the Built-in AI providers column. Source: calculated from other columns.
- Runs local models: Yes when it can use AI models running on your own computer, for example through Ollama. Source: the project's own pages, read by AI and checked against our rules.
- Any OpenAI-compatible service: Yes when you can point it at any service that speaks the OpenAI API by giving a URL and key. Source: the project's own pages, read by AI and checked against our rules.
- Any Anthropic-compatible service: Yes when you can point it at any service that speaks the Anthropic API by giving a URL and key. Source: the project's own pages, read by AI and checked against our rules.
- Uses coding-agent tools: Command-line coding agents, such as Claude Code or Codex, that it can use as its engine. Possible values: pi, OpenCode, Claude Code, Codex, Gemini CLI, Cursor CLI, Copilot CLI, Aider, Goose, Qwen Code, Kimi CLI. Source: the project's own pages, read by AI and checked against our rules.
- AI models included: Yes when the vendor itself provides access to AI models, free or paid, so you do not need your own key. Source: the project's own pages, read by AI and checked against our rules.
- Models offered by vendor: How many AI models the vendor's own cloud offers. Source: the project's own pages, read by AI and checked against our rules.
- Use your own keys: Yes when you can use your own API keys or AI subscriptions. Source: the project's own pages, read by AI and checked against our rules.
- Chat apps: The apps and channels you can talk to it through. Possible values: Telegram, Discord, Slack, WhatsApp, Signal, iMessage, Matrix, Email, SMS, Web, Terminal, Lark, DingTalk, WeChat, QQ, LINE, Mattermost, Microsoft Teams, Voice. Source: the project's own pages, read by AI and checked against our rules.
- Phone chat apps (calculated): Yes when the Chat apps column includes Telegram, WhatsApp, Signal, iMessage or SMS. Source: calculated from other columns.
- Desktop app: Yes when it has an installable desktop app. Source: the project's own pages, read by AI and checked against our rules.
- Conversation style: How conversations work, one of: One ongoing chat; Automatic sessions, shared memory; Sessions you manage, shared memory; Separate sessions, no shared memory. Possible values: One ongoing chat, Automatic sessions, shared memory, Sessions you manage, shared memory, Separate sessions, no shared memory. Source: the project's own pages, read by AI and checked against our rules.
- Shortens long chats itself: Yes when it summarizes or trims a conversation by itself once it gets too long. Source: the project's own pages, read by AI and checked against our rules.
- Memory stored as: Where its memory is kept, one or more of: Markdown files; SQLite database; Vector store; Knowledge graph. When the source says several are used but does not name them, the value is "More than one kind" instead of a guess. Possible values: Markdown files, SQLite database, Vector store, Knowledge graph, More than one kind, not named by the source. Source: the project's own pages, read by AI and checked against our rules.
- Memory shared by: Who shares one memory, one of: One session; One bot; One user; All bots. Possible values: One session, One bot, One user, All bots. Source: the project's own pages, read by AI and checked against our rules.
- Memory writer: Who writes to memory: only you, the agent when you ask, the agent on its own, or a background process. Possible values: Only you, Agent, when asked, Agent, on its own, Background process. Source: the project's own pages, read by AI and checked against our rules.
- Memory cleanup: How old memory is tidied: not at all, by summarizing, by merging duplicates, or by removing stale items. Possible values: Not at all, Summarize, Merge duplicates, Remove stale items. Source: the project's own pages, read by AI and checked against our rules.
- Memory cleanup every (hours): How often the memory cleanup runs, in hours. Unit: hours. Source: the project's own pages, read by AI and checked against our rules.
- Memory recall: How memory gets back into a conversation: all of it every time, keyword search, meaning-based search, or a tool the agent calls. Possible values: All of it, every time, Keyword search, Meaning-based search, Agent looks it up with a tool. Source: the project's own pages, read by AI and checked against our rules.
- Memory size limit (tokens): The maximum size of memory, in tokens. Unit: tokens. Source: the project's own pages, read by AI and checked against our rules.
- Memory kept when chat is shortened: Yes when memory survives after a long chat is summarized or trimmed. Source: the project's own pages, read by AI and checked against our rules.
- Memory kept after a fresh start: Yes when memory survives after you start a new chat. Source: the project's own pages, read by AI and checked against our rules.
- Keeps a profile of you: Yes when it stores facts about you in a separate profile. Source: the project's own pages, read by AI and checked against our rules.
- Keeps a project log: Yes when it keeps a running log for each project. Source: the project's own pages, read by AI and checked against our rules.
- Saves skills: Yes when it stores reusable skills or step-by-step procedures. Source: the project's own pages, read by AI and checked against our rules.
- Memory readable by you: Yes when memory is stored in a form you can read, such as text files. Source: the project's own pages, read by AI and checked against our rules.
- Memory editable by you: Yes when you can edit what it remembers. Source: the project's own pages, read by AI and checked against our rules.
- Memory exportable: Yes when you can export its memory to take it elsewhere. Source: the project's own pages, read by AI and checked against our rules.
- Memory history kept: Yes when earlier versions of memory are kept, for example in git. Source: the project's own pages, read by AI and checked against our rules.
- Writes its own memory (calculated): Yes when the agent or a background process writes memory, calculated from the Memory writer column. Source: calculated from other columns.
- Learns about you: Yes when it learns your preferences on its own over time. Source: the project's own pages, read by AI and checked against our rules.
- Writes its own skills: Yes when it creates new reusable skills by itself. Source: the project's own pages, read by AI and checked against our rules.
- Scheduled tasks: Yes when you can schedule tasks to run at set times. Source: the project's own pages, read by AI and checked against our rules.
- Schedules its own tasks: Yes when the agent can create scheduled tasks for itself. Source: the project's own pages, read by AI and checked against our rules.
- Helper agents: Yes when it can start helper agents that work in parallel or in the background. Source: the project's own pages, read by AI and checked against our rules.
- Agent teams: Yes when several named agents with different roles can work together. Source: the project's own pages, read by AI and checked against our rules.
- Messages you first: Yes when it can send you a message without being asked. Source: the project's own pages, read by AI and checked against our rules.
- MCP tools: Yes when it can use tools through the Model Context Protocol (MCP). Source: the project's own pages, read by AI and checked against our rules.
- Memory use when idle (MB): How much RAM it uses while idle, in megabytes. Unit: megabytes. Source: the project's own pages, read by AI and checked against our rules.
- Install size (MB): Disk space of the installed program, in megabytes. Unit: megabytes. Source: the project's own pages, read by AI and checked against our rules.
- Summary: A short description written from the project's own pages. Source: the project's own pages, read by AI and checked against our rules.
- Tags: Short topic keywords for search. Source: the project's own pages, read by AI and checked against our rules.
- Notes: Our editorial notes about this entry. Source: our editorial notes.