Overview
DCS Agentic lets you build automations and AI agents, run them, and prove every run with a signed, independently verifiable receipt — at a flat price, using your own API keys, with no credits and zero markup.
There are three ways to build, and one place to verify everything you run:
Connect apps with triggers and actions — "new email → summarise → save to Sheets". No code.
Drag-and-drop AI agents that reason and act, built from inputs, models and processors.
Plan and publish posts across every major platform from one calendar.
Every run is signed into a verifiable receipt chain with its real cost — auditable by anyone.
The left sidebar at a glance
| Tab | What it does |
|---|---|
| Home | Your dashboard — counts, recent activity, quick actions. |
| Flows | Visual Workflow Builder (triggers + actions). |
| Agents | Agent Studio — build AI agents from components. |
| Social | Social Scheduler + posting calendar. |
| Templates | Pre-built starters — the fastest way to begin. |
| Connect | Link your apps (Gmail, Slack, Notion…) via OAuth. |
| Trust | Verification & signed receipts for every run. |
| Keys | Your encrypted API-key vault (bring your own key). |
| Storage | Files and data your agents read and write. |
| Compute | Where heavier agent runs execute. |
Your first 5 minutes
Do these in order — connecting apps and keys first means everything's ready when you build.
Connect (integrations)
Linking apps is the first thing to do. Open Connect, choose an app, and authorise it. Once linked, that app appears as a step in Flows and as a component in Agents.
- Native apps: Gmail, Slack, Notion, Google Sheets, and many more.
- Any MCP server — plug in a whole ecosystem of tools instantly, no hand-built connector needed.
- Universal API connector — any REST/OAuth endpoint, using your own key.
If an app shows "mock", its OAuth credentials aren't configured on the backend yet — it'll connect for real once they're set.
Keys — bring your own key (BYOK)
Open Keys and paste your provider API key (e.g. OpenAI or Anthropic). Keys are encrypted at rest (AES-256-GCM) and never shown back to the browser or written to logs.
This is what makes DCS Agentic different on price: with your own key, the platform charge for a run is $0 — you pay the provider directly with no markup. Prefer not to manage keys? Use managed pass-through and we bill the exact provider cost, itemised to the cent.
Templates
The fastest way to start. Open Templates, pick a starter, and click Use template — it opens a pre-filled flow or agent you can tweak instead of building from a blank canvas.
Starters include: Research Agent, Social Media Agent, Sales Outreach, GitHub Code Review, Content Creation, Customer Support, DCS Rank Agent, and a Multi-Agent Orchestrator.
Flows — Workflow Builder
Use Flows for app-to-app automation that doesn't need AI reasoning.
Agents — Agent Studio
Build an AI agent by dragging components onto the canvas. The left Components panel has four groups:
| Group | Components | What they're for |
|---|---|---|
| Inputs | Text Input · File Upload · API Webhook | How data enters the agent. |
| AI Models | GPT · Claude · Gemini · DeepSeek · Your DCS Model | The "brain" that reasons. |
| Processors | LLM · Code Execution · Data Transform | Logic steps between input and output. |
| Integrations | Gmail, Slack… (whatever you connected) | Actions the agent takes in your apps. |
Build an agent, step by step
Start with the simplest chain — Text Input → LLM → one action — get it working, then add steps.
Social Scheduler
Open Social, compose a post, pick the platforms (X, LinkedIn, Instagram, Facebook, TikTok, YouTube), choose a date and time on the calendar, and schedule. It publishes automatically.
Live publishing needs each platform connected first in Connect.
Trust — verifiable runs unique to DCS
This is the feature no other automation platform has. After any run, open Trust to see its signed receipt: what ran, the real cost, and a cryptographic signature that you — or anyone — can independently verify, without trusting our servers. Receipts are chained, so a run's history is provable and replayable.
Cost ledger — see every cent
Every run shows the real provider cost with 0% DCS markup, itemised line by line. With BYOK your runs cost nothing on our side. There are no credits to run out of — check the ledger / transparency view after any run to see exactly what it cost.
Storage & Compute
Storage holds files your agents read from or write to. Compute is where heavier agent runs execute — there's usually nothing to configure to get started.
Troubleshooting
| You see | What it means | What to do |
|---|---|---|
| Badge "Operational" 🟢 | Backend healthy | Nothing to do. |
| A connector says "mock" | Its OAuth creds aren't set yet | Connect once credentials are configured. |
| Run shows no cost | BYOK = $0, or no model ran | Normal for BYOK; add a model step if you expected a cost. |
| Data missing after reload | Should never happen | Contact support — this is a bug, not expected. |
| "FREE" plan badge | You're on the free tier | Upgrade under View plans when ready. |
Plans
Flat base plus optional per-feature add-ons — no credits, ever. Hybrid $15/mo (core + one lane), per-feature +$10/mo each, Master $399/mo (everything). The only variable cost is third-party APIs, which you control via BYOK. See the pricing page for details.
Ready? Open the app and start from a template — Connect → Keys → Template → Test → Save → Trust.