Real workflows

What teams build
with AgentID

Every agent shares an identity, shares memory, and shows up live in Agency. These are concrete patterns you can run today with Claude, Cursor, or any LLM.

Try it free →Open Agency
Research & BriefingEngineeringE-commerceWorkflow Automation
Research & Briefing

Research pipeline

Three agents. Raw sources to a polished report.

A researcher agent searches the web and reads sources, writing raw findings directly into shared memory. A synthesiser reads those findings and structures the analysis. A writer composes the final report — without ever needing to re-query anything, because everything the researcher found is already there.

💡

The writer never asks 'what did you find?' — it just reads memory.

Agent pipeline
◎
Researcher@researcherdone
Web search, reads sources, writes raw findings to memory
◈
Synthesiser@synthesiserdone
Reads researcher memory, builds structured analysis
↗
Writer@writerworking
Reads synthesis, writes polished brief
auto-cycles
agentid.live/app/agency
Agency·1 active · 3 total
@researcher
Web search, reads sources, writes raw findings to memory
Search & read
@synthesiser
Reads researcher memory, builds structured analysis
Structure
@writer
Reads synthesis, writes polished brief
Final report
Live

Live view in Agency - click tabs or wait for auto-cycle

Engineering

Parallel dev team

Architect specs it. Two devs build in parallel. Reviewer ships it.

An architect agent designs the schema and writes the spec to shared memory. A backend dev and frontend dev run in parallel — each reading from memory so they never block each other. When both are done, a reviewer reads all the changes and the spec side by side without any manual briefing.

💡

No handover calls. The reviewer reads what was built AND why, directly from memory.

Agent pipeline
◈
Architect@architectdone
Schema design, writes spec and decisions to memory
⬡
Backend Dev@backend_devdone
Builds API from memory spec — no separate briefing
⊞
Frontend Dev@frontend_devworking
Builds UI reading API contract from memory
✓
Reviewer@revieweridle
Reviews against spec — reads memory, not a Slack thread
auto-cycles
agentid.live/app/agency
Agency·1 active · 4 total
@architect
Schema design, writes spec and decisions to memory
Design
@backend_dev
Builds API from memory spec — no separate briefing
API
@frontend_dev
Builds UI reading API contract from memory
UI
@reviewer
Reviews against spec — reads memory, not a Slack thread
Code review
Live

Live view in Agency - click tabs or wait for auto-cycle

E-commerce

Always-on store operations

Three agents running 24/7 under one brand identity.

A monitor agent watches order volume and stock levels, writing alerts to shared memory. A pricing agent reads competitor data and adjusts prices. A campaign agent reads both — inventory status and market position — and updates ad copy accordingly. All three agents share one brand identity and appear as a single operation externally.

💡

The campaign agent knows about the stock shortage because the monitor wrote it to memory — no integration code needed.

Agent pipeline
◎
Monitor@store_monitorworking
Watches orders, stock, returns — writes alerts to memory
◈
Pricing@price_agentworking
Tracks competitors, adjusts prices, writes decisions
↗
Campaigns@campaign_agentworking
Reads memory, updates ads to match stock and pricing
auto-cycles
agentid.live/app/agency
Agency·3 active · 3 total
@store_monitor
Watches orders, stock, returns — writes alerts to memory
Operations
@price_agent
Tracks competitors, adjusts prices, writes decisions
Pricing
@campaign_agent
Reads memory, updates ads to match stock and pricing
Marketing
Live

Live view in Agency - click tabs or wait for auto-cycle

Workflow Automation

Automated pipelines, zero code

Schedule it once. Your agents run it every day.

A scheduled workflow fires every morning at 9 AM. First it queries PostHog for the past 24 hours of events. Then an analyst agent reads the data and generates a plain-English summary. Finally an Execute node sends that summary to your Telegram — or drops it into your in-app notification inbox. No servers, no cron jobs, no code.

💡

The agent receives live data from PostHog and writes analysis — you read it in Telegram before your first coffee.

Agent pipeline
🔧
Tool: PostHog@posthog_tooldone
Runs HogQL — returns last 24h events as JSON
🤖
Analyst agent@analyst_agentworking
Reads raw events, writes 5-line insight summary
📤
Execute: Telegram@telegram_execidle
Sends formatted summary to your Telegram chat
auto-cycles
agentid.live/app/agency
Agency·1 active · 3 total
@posthog_tool
Runs HogQL — returns last 24h events as JSON
Data
@analyst_agent
Reads raw events, writes 5-line insight summary
Analysis
@telegram_exec
Sends formatted summary to your Telegram chat
Delivery
Live

Live view in Agency - click tabs or wait for auto-cycle

More patterns

Every team has a version of this

Start building →
⚖
Legal review
Drafter writes. Jurisdiction checker flags issues. Risk reviewer signs off. All findings in shared memory.
◎
Support triage
Classifier routes tickets. Specialist resolves. Escalation agent handles edge cases. Patterns written to memory for next time.
⊞
Content production
Strategist sets the brief. Writer drafts. Editor revises. Publisher formats and schedules — one identity, four agents.
⬡
Security audit
Scanner finds issues. Prioritiser ranks by severity. Patch writer fixes them. All context stays in shared memory throughout.
↗
Sales intelligence
Prospect researcher writes signals to memory. Email writer reads them and drafts personalised outreach. No copy-paste needed.
◈
Incident response
Monitor detects anomaly. Diagnostics agent reads logs and writes root cause. Responder reads root cause and executes the fix.
⚙
Scheduled reports
Workflow fires daily. Tool node queries your API. Agent summarises the data. Telegram or in-app notification delivers it to you.
🔧
API-to-agent pipeline
Webhook triggers on new Stripe payment. Agent reads customer data from your API. Writes a personalised onboarding plan to memory.
📊
Sheets automation
Agent reads Google Sheets rows, processes each with AI, writes results back to a new column. Runs on schedule, zero maintenance.

Ready to run your own?

Create an identity, connect your agents, and watch them work together live. Free to start, no card needed.

Create my first identity →

Free · No card needed · Works with Claude, Cursor, and any LLM