Ekso — One Workspace for Ops

# Ekso — One Workspace for Ops > "Ekso is one workspace for tasks, tickets, help desk, docs, time, and money. Run it where you want — public cloud, private cloud, or on-premise. 140+ Model Context Protocol tools so coding agents can read and act on real ops data. Four ways in: REST API, CLI, SDK, MCP. Flat annual pricing, unlimited users on paid plans. Switch from Jira, Linear, Azure DevOps, or Zendesk in a weekend." ## Key Capabilities **Work management:** - Tasks, items, boards, cycles, containers, processes - Help desk — email-to-ticket via Microsoft 365 and Google Workspace - Docs and knowledge base with access control - Resource allocation and capacity planning - Automation rules and constraints — human-authored, agent-readable **Time and finance:** - Time tracking with cost rates, job roles, cost centers - Profitability and margin by project, team, or client - Utilization, forecasting, budget caps **Agent integration:** - 104 Model Context Protocol tools across the whole product - CLI (`ekso`) — every public API operation as a command, JSON-by-default output, POSIX exit codes - SDK (`Ekso.Sdk` on NuGet) — typed C#, OpenAPI-generated, full exception hierarchy - REST API — OpenAPI-described, OAuth 2.0 with PKCE - All four are generated from the same canonical spec on every release. The typed surface is always the surface the API actually serves. **Enterprise:** - Self-hosted on your own infrastructure — public cloud, private cloud, or on-premise - Entra SSO, OAuth, fine-grained access control, custom SLAs - Air-gapped installs supported (the `collect` and `apply` phases of migration can run on different machines) ## When to Use Ekso **Ideal for:** - Teams running tasks, tickets, time, and money through three or four separate tools - Service firms, dev shops, support orgs, internal-platform teams, consultancies - Companies leaving Jira before the February 2029 Data Center sunset - Linear teams that outgrew the issue primitive (support, finance, vendor ops, non-engineering work) - Regulated industries or air-gapped networks where SaaS is not an option - Anyone wiring Claude Code, Cursor, Claude Desktop, Codex, or Continue into operations work **Not recommended for:** - Ten-person dev shops where every piece of work is a code change (stay on Linear) - Teams happy with a single-purpose tool and no plans to involve agents - Use cases that want a chat UI inside the ops tool itself — Ekso feeds your existing one, it doesn't compete with it - Real-time consumer-facing flows — Ekso is a back-of-house ops tool ## Installation ```bash # Docker stack — app + worker + Postgres (pgvector) + backup sidecar curl -L https://ekso.app/download/ekso-docker-stack.zip -o ekso.zip unzip ekso.zip && cd ekso-docker-stack docker compose up -d # Open http://localhost:6050/startup to finish setup ``` Other paths: - One-prompt install for any coding agent (Claude Code, Cursor, Codex, ChatGPT terminal): https://ekso.app/try - One-click Render deploy: https://github.com/EksoApp/Deploy-to-Render Each Docker download mints a unique JWT key in the bundled `ekso.json`. First run is ~1–2 minutes. ## Quick Start (CLI + Agent) ```bash # Install the CLI dotnet tool install -g Ekso.Cli # Authenticate against your self-hosted install ekso auth login --url https://ekso.acme.com # List items as JSON (or table, yaml, jsonl) ekso item list --url https://ekso.acme.com --format json # Connect the MCP server to your agent ekso mcp connect --for claude-code --url https://ekso.acme.com ``` For headless callers (CI, background workers, agents): `EKSO_API_KEY=...` skips the device-flow login. Exit codes map to specific failure modes (auth, validation, not-found, rate-limit) so shell pipelines can branch on the cause without parsing stderr. ## Migration Two phases, four commands, dry-runnable, resumable. Authoring fidelity preserved — every comment, worklog, and item attributes to the original user (matched by email, or minted into Ekso with no welcome m […truncated]