Somewhere between the third coding agent and the fifth model subscription, nobody can say where the tokens went. CPA Usage Keeper keeps the receipts for a CLIProxyAPI gateway, saving every request locally with scheduled backups so history survives restarts. The Overview page shows requests, success rate, tokens, cache hit rate, and estimated cost with sparklines and a calendar heatmap, and Realtime plots time to first token against total latency and shows rolling cache levels and model speed. Analysis breaks usage down by API key, model, credential, or provider with donut charts and ranked lists. The Request Events log filters by model, source, and status, adjusts its columns, and exports records. Credentials cards show success rates, remaining five-hour and weekly quotas, priorities, and enable switches, and quota history charts how many tokens and dollars each percentage point costs over a cycle. Individual API keys can be given a read-only usage view. Deployment uses a Docker Compose file that bundles CPA with Keeper or runs Keeper beside an existing gateway, and the interface switches between English and Chinese and between light and dark themes. Check the dashboard before the invoice arrives and see exactly who spent what. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Headroom answers the question every multi-plan AI subscriber asks mid-task: which of my accounts still has room to work? It puts the live limits, reset times, credits, and spend of Claude, Codex, Cursor, Copilot, Grok, Antigravity, and Vercel AI Gateway accounts on one self-hosted dashboard instead of seven provider pages. Connect each account once, through its official CLI sign-in or a credential import, and a background scheduler refreshes every figure on schedule, renews tokens before they expire, and keeps 90 days of readings. Five-hour sessions, weekly pools, per-model limits, credit balances, banked resets, and on-demand spend sit side by side, and unreported figures show as unknown rather than a fake zero. Compare ranks same-provider accounts by the room they have left and names the one to use next, while Timeline lays every reset window on a two-week calendar. Analytics charts when you work, when you run dry, and whether each plan fits, and Wallet totals subscriptions, usage spend, and top-ups in one currency with renewals and budget alerts. In-app notices, Telegram messages, and signed webhooks warn about low limits, expiring resets, and broken sign-ins, and optional automations can spend a banked Codex reset or log a credit top-up the moment it lands. Personal, work, and side-project accounts stay separate under one owner login with passkeys and a Privacy Mode that blurs emails on screen. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Software developers running Claude Code, Cursor, Codex, and Aider use AgentsView to trace token expenditures, search full session transcripts, and analyze multi-agent concurrency timelines. Engineers can search through historical developer sessions to locate specific code modifications, prompt chains, or debugging rationales across extensive programming workflows. The real-time activity timeline visualizes concurrent agent operations, revealing peak execution windows, task run durations, and automated tool invocations. The usage analytics engine calculates financial costs and token volumes per model by applying live pricing tables while accounting for prompt caching read discounts. Developers inspect granular turn-by-turn session transcripts complete with raw model payloads, file difference diffs, executed shell commands, and syntax errors. The recall browser extracts key technical context from past programming sessions, organizing facts into an indexed reference knowledge corpus for reuse across future project tasks. Administrators can replicate SQLite session tables into DuckDB analytics files or PostgreSQL warehouses to query longitudinal metrics using SQL analytical tools or custom reporting scripts. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Claude Code Agent Monitor puts every AI coding agent session on one live control room, so you can see what each agent is doing, what it costs, and where it got stuck. Claude Code, Cursor, and Codex sessions land together in a dashboard, a sessions list, an activity feed, and a kanban board. Open any session to read the full rendered conversation with highlighted code, tool calls, command output, and sub-agent timelines. The analytics page charts spend by model and weekday, token totals, cache hit rates, and a 52-week activity heatmap. Workflow views map parent and child agents, delegation, compactions, errors, and concurrency. An Agent Config explorer lists your skills, subagents, slash commands, hooks, plugins, MCP servers, and memory files in one place. The Run Agent page starts a conversation or one-shot job, streams output, and resumes earlier sessions without leaving the browser. Webhook alerts, push notifications, and a REST API keep the rest of your tooling informed. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Watch every coding agent you run, on every machine, from one live browser dashboard. Sessions appear as cards, a sortable list, or a graph, grouped by machine or project, each showing its plan and current step, subagents, model and effort, context fill, git branch, and cost. Anything waiting on you, like a permission, a question, or a plan to approve, rises to a Needs you panel with the longest wait first, and you can answer it from the page or in the terminal. A Problems panel flags sessions that are stuck, failed, or colliding on the same file. The Plans board sorts every plan step into pending, in progress, review, blocked, and done. A session page adds a timeline of loops and subagents, changed files, and token and context charts. Spend is broken down by model and machine, and plan windows show how close each provider is to its limit. History keeps ended sessions by day, a wall view suits a spare screen, and browser alerts fire when a session needs 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.

OpenClaw Dashboard is a secure real-time monitoring dashboard for OpenClaw AI agents that tracks live sessions, API costs, rate limits, and host health from one browser page. The overview refreshes every five seconds with live session activity, token totals, and lifetime spending, so you can see what every agent is doing right now. Cost views break spending down by model, session, and time period, and rate limit panels track Claude and Gemini usage against rolling windows with browser alerts before quotas run out. A live feed streams agent messages across all sessions, and a session browser adds search, status filters, timelines, and a 30 day activity heatmap. You can read agent memory files (MEMORY.md, HEARTBEAT.md, daily notes) and edit workspace files, skills, and configuration with JSON validation and automatic backups. Host panels report CPU, memory, disk, and temperature with 24 hour sparklines, plus Docker containers, cron jobs, real-time logs, and one-click service restarts. A security page shows firewall rules, open ports, fail2ban status, and SSH login history, and sign-in uses PBKDF2 hashed passwords with optional TOTP two-factor codes. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

Oh My CPA gives a CLIProxyAPI gateway a proper control room, where providers, OAuth credentials, client keys, model routing, and the cost of every request are managed and watched from one self-hosted console. You add providers such as Codex, Claude, Gemini, xAI, DeepSeek, Vertex AI, and OpenAI-compatible services, each with credentials, retry and error policies, priority, weight, and an enable switch the gateway actually enforces. OAuth accounts are signed in from the console, with quota window readings, refresh schedules, and per-credential model aliases kept next to the keys they belong to. Client keys can be created, named, and revoked, and those names follow the traffic into request records and filters so you always know which key spent what. The dashboard follows request volume, token throughput, cache hit rate, and cost from the last fifteen minutes out to ninety days, with a year-long token heatmap for the long view. Insights compare keys, models, providers, and credentials against the previous period, while the live routing view shows which credential each request travelled to and whether it was rerouted. Every request stays searchable, with duration, time to first token, token breakdown, and the raw log in a detail drawer. A playground tests any routed model with text and images before your users do. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

CPA Manager Plus is a self-hosted observability dashboard and management panel that tracks every AI request flowing through your CLI Proxy API gateway, breaking down failures, costs, and account health across providers like OpenAI, Anthropic, xAI, and Codex in one interface. When a request fails, drill into the persistent history to see status codes, affected models, latency, and redacted failure evidence without exposing raw response bodies. The cost analytics engine breaks down token consumption and estimated spend by model, provider, account, API key, project, channel, and time range while tracking input, output, reasoning, cache, and service-tier pricing semantics separately. Model prices sync automatically from models.dev with LiteLLM and OpenRouter fallbacks, and you can add local overrides for aliases or internal models. For teams running Codex or xAI accounts, the health inspector reads quota windows, reset evidence, credential state, and workspace status on a configurable schedule, routing credential failures into an action queue for review rather than letting them silently degrade throughput. Deploy the Lightweight Panel to replace your existing CPA management UI without adding another service, or run Full Mode as a single Docker container that adds the Manager Server with persistent SQLite storage for request history, historical analytics, and automated account inspections. Export or import request history as JSONL for external analysis, and back up the SQLite files alongside your encrypted management keys. 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.

Your AI application is burning through API tokens faster than you can refresh the billing page, and you have no idea which prompt template is responsible. OpenLIT plugs that visibility gap with a self-hosted observability platform built specifically for LLM workloads. Add one line of code to instrument 90+ LLM providers, agent frameworks, and vector databases, then watch every request flow through a tracing dashboard that shows tokens consumed, latency measured, and dollars spent per call, per model, per environment. The requests view lists every LLM interaction with provider, model, cost, and token breakdown in a filterable table, while the trace detail panel lets you drill into individual spans to read the exact prompt sent and response received. Prompt Hub turns prompts into versioned artifacts you deploy, rollback, and A/B test without touching application code. OpenGround compares models side by side on the same input, so you can evaluate cost-versus-quality tradeoffs before committing to a provider. Automated evaluations run LLM-as-a-judge scoring on live production traces, flagging hallucinations, bias, and toxicity in real time. The Vault stores and rotates API keys centrally so secrets stay out of your codebase. Custom dashboards let you build drag-and-drop monitoring views with charts, stat cards, and tables backed by SQL queries against ClickHouse. GPU utilization, memory, temperature, and power metrics feed into the same platform for end-to-end infrastructure visibility. 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.

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.

Backed by Y Combinator (S24) and processing traces from thousands of AI agents in production, Laminar is the open-source observability platform that treats agent debugging as a first-class engineering discipline rather than an afterthought. Its OpenTelemetry-native SDK auto-instruments Vercel AI SDK, LangChain, OpenAI, Anthropic, Gemini, Browser Use, Stagehand, Mastra, Pydantic AI, and the OpenAI Agents SDK with a single line of code, capturing every LLM turn, tool call, and sub-agent delegation as nested spans with full input/output data and token costs. The Signals engine lets you describe failures in plain language — "agent is stuck in a loop" or "tool returned empty results" — then reads every trace and alerts via Slack when it detects a match. A built-in debugger records runs and replays them from cache so each iteration takes seconds, designed for Claude Code, Cursor, or Codex to drive the repair loop via the MCP server or CLI. Run code-first evaluations in Python or TypeScript locally or in CI/CD pipelines, build datasets from production traces, and query everything with raw SQL through custom dashboards, the in-app editor, or your coding agent. The Rust backend delivers 20x trace compression, a custom real-time streaming engine, ultra-fast full-text search, and gRPC ingestion, while ClickHouse powers columnar analytics and PostgreSQL stores application state. 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.

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.

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.

A looping agent burns money quietly until someone can see where it drifted, and PandaProbe is an observability and evaluation platform for AI agents that makes that drift visible step by step. Capture complete executions as sessions, traces, and spans spanning model calls, tool calls, multi-agent handoffs, and custom logic, then follow a looping or failing run back to the decision where it drifted. Score traces and sessions with agent-specific metrics, run asynchronous and sampled evaluations so judge models do not cost more than inference, and use lightweight trajectory signals to catch stalls, bad tool use, and regressions over time. Bring agents built with LangGraph, CrewAI, the Claude Agent SDK, the OpenAI Agents SDK, and other integrations, with MCP tool calls captured automatically where supported. Organize work by user, organization, and project, send data through an SDK or HTTP API with project keys, and query filtered, paginated traces when you need the raw record. Keep sensitive workflows local when that matters: prompts, traces, and evaluation results stay inside the deployment you run, instead of leaving for a hosted console. 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.

Agenta delivers a comprehensive open-source LLMOps workspace that covers the full lifecycle of AI application development — from prompt engineering through production monitoring. The platform supports 15+ model providers including OpenAI, Anthropic, Google Gemini, Mistral, Groq, Together AI, Azure, AWS Bedrock, and self-hosted models via Ollama, enabling teams to switch between providers without code changes. The prompt playground allows side-by-side comparison of different configurations, while the evaluation system offers LLM-as-a-Judge assessment, 20+ pre-built evaluators covering semantic similarity, regex matching, and factual accuracy, plus custom Python evaluators for domain-specific requirements. Teams run evaluations through both the web UI for subject matter experts and the Evaluation SDK for programmatic CI/CD integration. The observability layer captures full trace visibility across complex agentic workflows, flagging quality issues like hallucinations and off-topic responses in real time. Human annotation workflows let domain experts review and annotate LLM outputs, feeding corrections back into the evaluation loop. The architecture supports Chain of Prompts, RAG pipelines, and multi-step agent workflows, integrating with frameworks like LangChain and LlamaIndex. Self-hosting deploys via Docker Compose with Traefik for routing, requiring only a clone, environment configuration, and a single docker compose command. On RepoCloud, deploy Agenta on a dedicated VPS with root SSH access, persistent storage for evaluation datasets and traces, and complete control over model provider credentials, all under the MIT license with no usage restrictions.

Evidently gives your models and LLM apps a proper check-up and keeps every result on one shared dashboard. Project dashboards chart predictions against targets, data drift, and quality scores over time, with counters for model calls and the share of drifted features. A reports page lists every evaluation snapshot with tags, metadata, timestamps, and view or download buttons, and a search box filters them by metadata. Over 100 built-in metrics cover data quality, classification, regression, ranking, recommendations, text descriptors, and LLM judges. Test suites add pass or fail conditions, so a nightly job can flag a drifting feature before customers notice a bad prediction. Datasets, traces, and prompts have their own tabs, which keeps retrieval and generation experiments beside the scores they produced. Reports export as HTML or JSON for sharing outside the dashboard. Results live in a local folder, a SQL database, or S3-compatible storage, and a remote workspace accepts snapshots sent from jobs running anywhere. Built-in demo projects with bike demand forecasting and e-commerce review classification show how a finished dashboard looks before you wire in your own data. 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.