A control room for AI agents, where every run is recorded, capped, and ready to resume after a crash. Mission Control opens first with agent and run counts, tokens, success rate, and cost per day. A Templates page applies ready-made agents for code review, KYB due diligence, lead scoring, market monitoring, and release notes in a single screen. Every run gets a Report tab with the finished answer and a Timeline tab with each message, tool call, and charge. Tools are set to allow, ask, or deny. Spend caps pause a run instead of failing it. An agent can stop and ask a person a question. The Playground starts a run from a form and shows the same request as curl, so each agent doubles as an API. Triggers start capped runs on a schedule, from a signed webhook, or on a GitHub pull request. Learning mode keeps notes, skills, and workflows from each job so repeat work needs fewer tokens. Automated code review, customer research, and competitor digests are typical recurring jobs to hand it. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Agents that keep working after you close the tab: a13n gives a team one shared place to define them, run them, and watch what they did. In the Console, an administrator adds models from OpenAI, Anthropic, Gemini, Bedrock, Azure, OpenRouter, DeepSeek, Ollama, and others, then creates agents with instructions, skills, memories, connections, browsers, and sandbox environments. A Try agent button opens a conversation, and the Debug view lays out every run step by step, with model calls, tool calls, token counts, cache rates, and durations. Sessions and schedules keep a history of conversations and recurring jobs. The Usage page charts spend, runs, requests, and tokens by agent, model, or member across a time range. Findings analyzes traces, flags answer quality and tool failures, and lets reviewers confirm or dismiss each one with a suggested improvement. Each agent has a Connect to app screen with its ID, service URL, and API key, so applications can call it over an HTTP API or SDK. Workers lease runs from PostgreSQL and save checkpoints to object storage, so interrupted work resumes. Customer support teams, internal platform groups, and product builders use it to ship agent features without writing their own orchestration layer. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Run a whole fleet of AI agents as managed services, each with its own model, memory, schedules, and sandboxed workspace, from one admin console. Create an agent, point it at OpenAI, Anthropic, OpenRouter, or any OpenAI-compatible endpoint, store its API keys encrypted, and it is ready to chat in the browser. People sign in with a device key and get a session list, streaming replies, and file attachments. The Agents table shows status, last activity, sessions, errors, and token spend for the past 24 hours, and the Stats page tracks users, sessions, events, and total tokens. Cron or one-shot schedules let an agent send a daily brief, and each run is logged. Owner, user, and guest roles control who may talk to an agent or change it, and policy and approval tabs gate risky tool calls. Skills and MCP servers are registered once and enabled for one agent or the whole fleet, and instructions can be pushed to every agent at once. Sessions, memories, and secrets sit in PostgreSQL, so state persists across gateway restarts. Use it for an internal agent platform, a support desk, or one agent per customer. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Edict is a multi-agent orchestrator that runs every AI request through a court-style chain: a crown prince agent triages it, a planning department drafts the work, a review department can veto the plan, and specialist ministries execute what survives. Casual chat gets a direct reply, while real commands become kanban tickets whose titles are scrubbed of file paths and metadata. A vetoed plan returns for rework; an approved one is dispatched in parallel to ministries for finance, documentation, engineering, compliance, infrastructure, and HR. Commands arrive through Feishu, Telegram, or Signal channels configured in OpenClaw. The dashboard shows cards by state with heartbeat badges and stop, cancel, and resume controls, plus a department monitor and a memorial archive that turns finished work into five-phase timelines you can copy as Markdown. Nine edict templates add parameter forms and cost estimates, an officials panel ranks agents by merit score while tracking token spend, and per-agent model switching restarts the gateway in about five seconds. Agent thinking, tool calls, and results stream live. A permission matrix limits who may message whom, illegal state jumps are rejected, and every transition lands in an audit log. Worked examples cover competitive analysis, code review, and weekly reports. Dashboard login protects the console, and a systemd unit keeps it running across reboots. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Cotal puts your AI agents in one shared room: Claude Code, Codex, and OpenCode discover each other, message channels, hand off work, and ask you to approve the result. Post to a channel and every subscriber hears it; message one peer and the durable inbox holds it until they are free; address a role like reviewer and one available reviewer claims the job. Presence shows who is actually there, and attention modes decide which channel traffic may interrupt an agent mid-task. A manager starts and stops agent processes in tmux or cmux, here or on remote hosts you authorize, without sitting between their messages. For work that waits and branches, Cotal Lang programs launch agents, pause at human checkpoints with deadlines, and resume from their recorded steps after an interruption. Services join the same space by publishing command contracts that authorized agents can discover and invoke; the bundled Linear adapter shows the pattern. Supervise from a terminal console with the live roster and activity feed, or from the read-only browser dashboard and its graph of agents and channels. Spaces stay separate, every message carries a checked identity, and per-agent grants decide who may delegate, spawn teammates, or call a service. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Describe the outcome once, in plain language, and a small crew of AI agents divides the work between them and reports back in the same chat. You type the request in the team chat; Boo Zero, the team lead, splits it into jobs on a shared board and hands each card to the teammate best placed to do it. Teammates run on Claude Code, Codex, Hermes, OpenClaw, or Clawboo's built-in runtime, and one team can mix them freely, so the researcher, the writer, and the coder can each work on their strongest tool. You follow every card live on the board and in the chat, watch the team graph to see who is doing what, and add a reviewer agent when a second pair of eyes matters. Finished work returns to Boo Zero, which pulls the pieces together and answers you in the chat. Ready-made teams from the marketplace cover marketing squads, dev teams, and research crews, and you can bring an API key from Anthropic, OpenAI, OpenRouter and more, a free local model through Ollama, or your ChatGPT subscription through the Codex CLI. Spend is tracked per team with budget warnings before costs surprise you. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Instead of one overworked chat box, Captain Claw gives you a Flight Deck for a whole fleet of AI agents: spawn specialist teams, watch each agent work live, and chat with any of them from a single dashboard. Describe a goal and Agent Forge designs the team, with roles, tools, and operating procedures, ready to spawn in one click. Agent Council runs moderated debates that end in a vote and exportable minutes, Basna sends the same question to blind specialists and merges their answers by learned reliability, and Vatra lets a team co-write a deliverable on a shared blackboard. Code takes a request through a plan you approve, a build, and three independent reviewers, committing every phase to git. Bat keeps working until an independent judge says the job is done, and Flows run deterministic automations on triggers. Mix OpenAI, Claude, Gemini, DeepSeek, OpenRouter, or local Ollama models per agent, with six memory layers shared across sessions and token and cost meters for every agent. Teams get admin-managed users and shared agents, and agents reach you through WhatsApp, Telegram, Slack, or Gmail while Claude can drive the fleet over MCP. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Prototype multi-agent AI systems without writing orchestration code: AutoGen Studio is Microsoft's low-code interface over the AutoGen AgentChat framework. You compose teams of LLM-powered agents in a visual Team Builder, either by drag-and-drop from a component library or by editing the declarative JSON specification directly. Each agent gets a model, a prompt, tools (Python functions), and the team gets termination conditions and an orchestration pattern, sequential or LLM-driven. The Playground runs teams interactively with live message streaming between agents, a visual control-transition graph, tool-call and code-execution tracking, and pause/stop controls, which makes it a practical debugger for agent behavior. Finished teams export as JSON for use in any Python application via the TeamManager class, or serve as an API endpoint. Any OpenAI-compatible model endpoint works, including local servers like Ollama or vLLM. Microsoft labels it a research prototype: use it for prototyping and evaluation, and build production systems on the underlying AutoGen framework.

SandBase Harness gives your AI agents a proper home with sessions, tools, sandboxes, and an audit trail instead of a pile of scripts. The console lists agents with their model, status, and version, and an agent page shows its system prompt, attached tools, and skills. Every session keeps a resumable event log, and a transcript view opens any user or agent message as a rendered or raw event. Built-in tools run in a local process, per-session Docker containers, Kubernetes, or a self-hosted worker queue, and permission policies can always allow a tool or always ask a human first. MCP toolsets, credential vaults, and skill packages extend what an agent can reach, with Anthropic's xlsx, pptx, pdf, and docx skills listed next to your own. Memory stores, webhooks, scheduled deployments, and outcome definitions cover longer-running work. Settings include a model provider form, API keys for bearer authentication, and an interactive API reference for sessions, agents, environments, and vaults. A TypeScript SDK and a bundled MCP bridge connect other tools to the same runtime. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache 2.0 licensed.

When an AI agent breaks in production, Tracely freezes that exact run into a replayable regression test and blocks the pull request that would ship the bug again. Span soup hides the failure. OTLP/HTTP traces group into conversation threads, and a waterfall marks the failing span in red with its input and output. Nobody knows if an answer was good. Evaluators sit as columns on the trace table, mixing model-free checks with LLM judges at conversation, run, or span level. Thirty-one broken runs, one issue. Clustering turns them into a single counted issue and can suggest the evaluator that would have caught it. A fix nobody can prove. One click promotes a failing trace into a hermetic case with recorded fixtures and a fail-to-pass contract, and scenarios send a red-team model after an adversarial goal. Model bills on every CI run. The suite replays offline with no API keys and no model spend, and tracely gate posts a commit status, comments on the pull request, and blocks regressions. No warning when it breaks again. Alert flows drawn on a canvas react to failed gates, broken conversations, unseen failure modes, or rising failure rates by messaging Slack, sending email, calling a webhook, or prompting an LLM. Coding agents can read traces and write evaluators through an MCP endpoint. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Before an AI agent meets real users, put it through a security checkup it cannot sweet-talk. DodoGuard is a web console and command line where teams register their agents, run evaluation and red-team tasks against them, and decide if each one is cleared for launch. The overview page shows a security index, open risks, remediation counts, and a pipeline that runs from asset intake through detection and remediation to launch monitoring and audit reports. An asset ledger lists every managed agent with its platform, environment, and risk level. A guided wizard creates assessment tasks that pick a test library for prompt injection, instruction leakage, tool misuse, and output risks, then runs black-box conversations against the agent. Evaluation configs cover prompts, providers, variable cases, and several assertion types, with OpenAI, Anthropic, Azure, Bedrock, Ollama, and any OpenAI-compatible endpoint supported. Findings land in a closed-loop board where owners confirm, fix, retest, and close each risk, and finished tasks export as reports for release evidence. Docker Compose with SQLite brings up the web UI on port 3000, and the same engine runs in continuous integration. Security teams use it for launch gates, platform owners use it for regression checks after prompt changes, and auditors use it for evidence. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Preloop is a self-hostable control plane that governs every AI agent tool call and model call, and can run ticket-triggered agent work that returns as reviewable pull requests. An MCP firewall can allow, deny, or hold any tool call for approval, with require-justification rules written as YAML and CEL policy. Human approvals arrive on mobile, watch, Slack, Mattermost, email, webhook, or the command line, and agents can ask a person a question before continuing. The model gateway speaks OpenAI, Anthropic, and Gemini formats, applies budgets and allowed-model lists, and attributes token spend by model, agent, session, API key, flow, and user. Flows start agents from GitHub, GitLab, Bitbucket Cloud, Jira, webhooks, or CI jobs, in hosted containers or on private runners that connect over an outbound WebSocket. Twenty-one presets review pull requests, implement issues, triage tickets, scan for vulnerabilities, and collect evidence packs. Operators can steer a running agent from the console, phone, or watch with notes, then review the resulting pull request. One session timeline records tool calls, model calls, policy decisions, approvals, spend, outcomes, and synchronized recordings. Fleet Map shows people, devices, agents, models, and observed work in one view. An account kill switch blocks gateway and tool traffic and stops managed runs, with audited recovery. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

To pull a team's skills, MCP servers, hooks, prompts, and agent files out of scattered Git repositories, Observal keeps them in one reviewed registry and traces real sessions to show which ones help. Build an agent visually, with a live config preview for every harness. Submissions wait in an admin review queue, where reviewers inspect side-by-side version diffs before approving a release. Teamspaces give each team its own namespace, and approved releases can be forked into independent drafts. One pull command then writes the right config for Claude Code, Cursor, Kiro, Codex, Copilot, Pi, OpenCode, Antigravity, or Goose. Every approved resource is discoverable through the open Agentic Resource Discovery standard, and a running agent can hand part of its task to another approved agent or a remote Agent2Agent service. Hooks record each session as a trace of prompts, tool calls, MCP round trips, tokens, and errors, with turn-by-turn replay for debugging and audits. Insight reports, generated through providers such as Anthropic, OpenAI, or Ollama, list what is working, what is hindering, and quick wins. A dashboard follows adoption, cost, and department usage, a leaderboard ranks agents by downloads, and audit logs, SAML single sign-on, and SCIM provisioning are included. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Every call your AI agents make becomes a searchable trace in LangWatch, and the same platform tests and scores those agents before a new prompt or model reaches users. The Trace Explorer filters by origin, user, model, cost, and evaluator, reads each run as a waterfall, flame graph, topology, or sequence diagram, and takes traces from the Python, TypeScript, and Go SDKs or any OpenTelemetry source. Agent Testing sends simulated users, in text or beta voice calls, through your agent while a judge passes or fails every criterion with its reasoning, adds multi-turn red teaming, and compares two agents side by side. The Experiments workbench scores prompts and agents on datasets with LLM judges, exact match, and RAG checks, then compares runs on cost, latency, and pass rate, while online evaluations monitor live traffic and guardrails block unsafe replies. Prompts keep version history and staging or production tags, Workflows wire prompts, Python code, and HTTP calls on a visual canvas, and annotation queues route traces to human reviewers. Alerts fire when a graph crosses a threshold, automations post to Slack, email, or webhooks, and scheduled reports go out by email or Slack. Claude Code, Codex, Copilot, Gemini CLI, and opencode report session cost, pull request cost, and terminal replays, and an MCP server exposes trace search and prompt management to coding assistants. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

When an AI agent goes wrong at step seven of twelve, Opik, an open-source LLM observability platform, shows the exact trace, tool call, and prompt behind it, then scores your fix against test datasets. Full trace trees capture every LLM call, retrieval step, and tool execution, and you can annotate traces and spans with feedback scores in the UI or through the SDK. Native integrations for OpenAI, Anthropic, LangChain, LlamaIndex, CrewAI, Google ADK, and DSPy, plus OpenTelemetry for any language with an OpenTelemetry SDK, feed traces in, or add the @track decorator to your own functions. Datasets and experiments compare prompt and model versions, scored by LLM-as-a-judge metrics for hallucination, moderation, and RAG relevance, with a PyTest integration to run evaluations in CI. A Prompt Playground tests prompts across models, and the Agent Optimizer SDK helps improve prompts and agents. Production dashboards track feedback scores, trace counts, and token usage, online evaluation rules score live traces, and optional guardrails add safety checks. An MCP server lets Claude Code, Cursor, VS Code Copilot, Codex, or opencode read traces and run evaluations from chat. Run it yourself with the ./opik.sh script on Docker Compose, or a Helm chart on Kubernetes, so trace data stays in your infrastructure. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

Promptfoo replaces vibes-based prompt tweaking with measurable tests, then attacks your AI app the way a hostile user would. An evaluation matrix lines up every prompt and model side by side, with pass or fail badges, assertion messages, latency, token counts, and cost in each cell. A red team wizard walks through the target, application details, plugins, and strategies, then probes for prompt injection, data exposure, broken access control, and harmful content. Preset bundles map plugins to the OWASP LLM Top 10, NIST AI RMF, MITRE ATLAS, GDPR, ISO 42001, and the EU AI Act. Meta Agent and Hydra Multi-Turn strategies adapt attacks across a whole conversation instead of firing one-shot prompts. Providers span OpenAI, Anthropic, Gemini, Bedrock, Azure, and local Ollama models, and any HTTP endpoint, Python script, or JavaScript function can be tested as a target. The history page lists every run with pass rates and raw scores, and prompt and dataset pages show which templates and test sets each run used. A provider allowlist file narrows the models offered in the web editor for cost control. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Production AI agents fail in quiet ways, and Latitude records every trace, session, and tool call so each failure can be found, grouped, and fixed. One line of telemetry from the TypeScript or Python SDK, or any OpenTelemetry exporter, sends traces from OpenAI, Anthropic, LangChain, LlamaIndex, CrewAI, and dozens of other providers and frameworks. Failing traces are clustered into signals with status, size, and trend, and flaggers and annotation queues let reviewers score outputs by hand. Evaluations are generated from those judgments and tracked with an alignment score. Semantic and exact-text search covers every trace, and saved searches, filters, and behavior groupings keep recurring questions one click away. Dashboards report cost per session, tool failure rates, and agent score across outcome, reliability, speed, and safety. Memory views show how an agent's stored memory changes over time, with diffs and the trace behind each edit. Monitors raise alerts, regression datasets replay fixes against the traces that failed, and Agent Dispatch hands the evidence to a coding agent. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Backed by Y Combinator and trusted by over 2,300 companies processing billions of observations monthly, Langfuse is the most widely adopted open-source platform for building, monitoring, evaluating, and debugging LLM applications. The hierarchical tracing engine captures every LLM call, tool invocation, retrieval step, and agent action as nested spans based on OpenTelemetry, with automatic cost calculation, latency tracking, and token usage attribution across sessions and users. Prompt Management separates prompts from code with versioned artifacts, label-based deployments, one-click rollbacks, and runtime SDK fetching with server-side caching, while linking every generation back to its exact prompt version for attribution analytics. The evaluation system supports LLM-as-a-judge scoring, heuristic code evaluators, user feedback collection, and manual annotation workflows that run automatically on production traces or against curated datasets. The Playground enables interactive prompt testing on real production inputs with side-by-side model comparison across providers. Datasets and Experiments define test cases for systematic benchmarking with comparative result visualization. Native SDKs for Python and TypeScript provide decorator-based instrumentation, while 100+ integrations cover LangChain, LlamaIndex, OpenAI SDK, LiteLLM, Vercel AI SDK, and any OpenTelemetry-instrumented framework. The analytics dashboard surfaces cost breakdowns, quality scores, latency percentiles, and usage trends across models and prompt versions. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

One endpoint for every MCP server you run: MCPHub gathers local and remote Model Context Protocol servers behind stable URLs, so AI clients connect once and see what you allow. Claude Code, Cursor, and Cherry Studio can use four routes: everything at /mcp, a named group at /mcp/{group}, one server at /mcp/{server}, or semantic tool search at /mcp/$smart. Smart routing runs vector search on PostgreSQL with pgvector and an OpenAI-compatible key, so the model gets a shorter tool list. Servers attach over stdio, SSE, or streamable HTTP, OpenAPI services can be wrapped as MCP tools, and changes apply without downtime. In the other direction, MCP tools can be exported as an OpenAPI 3.0.3 spec for OpenWebUI and other OpenAPI clients. Bearer keys, OAuth 2.0 (client and server modes), and GitHub or Google sign-in in database mode control access, with per-user visibility for servers and groups. Encrypted per-user credentials let one shared server use each person's own key. Logs show tool-call activity, request status, and latency, health checks track each connection, and the dashboard controls which tools, prompts, and resources each server exposes. Large text tool results can be compressed, interactive MCP Apps are proxied through MCPHub routes, and a built-in market lets you browse a server catalog. Typical uses include a team-wide gateway, per-project groups, and direct single-server routes for debugging. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. Apache-2.0 licensed.

MCPProxy puts one safe endpoint in front of every MCP server your AI agents use, and it holds each new server in quarantine until a person approves it. Add stdio or HTTP servers on the Servers page, then review the captured tool definitions and scanner findings in the Review queue before any agent can call them. The Tools page lists every tool with a Read, Write, or Destructive tier, an approval state, and a usage count, plus switches to hide whatever you do not want exposed. Agents see a short set of discovery and call tools instead of the full catalog, and the Home dashboard estimates how much context that saves per request. The Activity Log records each call with its server, intent, status, and duration, and exports for audits. The Usage page charts calls per tool, response sizes, errors, latency, and activity over time, filtered by profile, client, or token. Profiles and agent tokens scope what each client may reach, a Secrets page keeps credentials out of config files, and a setup wizard connects Claude Code, Claude Desktop, Cursor, Windsurf, and VS Code. Platform teams and solo builders get one place to see and control everything their agents can touch. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.