Specialist agents that do the job — and stop for your approval.
Open-source agent orchestration from Workstation. Describe a job in plain English; the agent plans it, signs into your real tools, streams every step, and pauses before anything is sent, spent or deleted. Self-host it, or run it as a managed service.
Most AI tools give you text back. Jobshout gives you a completed job: a reply drafted in your mailbox, a pull request reviewed, a security report with reproduction steps, an SEO score with fix prompts, a researched brief with citations you can open. The agent does the work; you decide what ships.
在 YouTube 上观看: https://youtu.be/ACmTPIoH7Mc
Live analyse of workstation.co.uk — SEO score 44 with title, meta and H1 fix prompts.
Brief an agent the way you would a colleague — one sentence is usually enough.
Built-in agents each with their own tools, prompt and rules — Mail, Career, Research, Articles, Security, PR review, Images, SEO Analyst and more.
Agents sign into Gmail, GitHub, job boards and live sites, then click through them like a person would.
Mark which tools need a human decision. Those calls pause the run until you approve or reject.
01
Type it in chat, or fill the short form the agent publishes. "Chase the two overdue invoices." "Review pull request 184." "Analyse this site for SEO."
02
It picks the tools it needs and asks for anything missing — a target URL, a repository, the sender to watch. It will not invent details it was not given.
03
Signed into your tools, one step at a time: search, open, read, cross-check, draft. Each step streams into the run as it happens.
04
Reach a gated tool and the run pauses mid-loop. State is saved; it resumes from that exact step when you approve — or takes your reason when you reject.
05
A draft, report, brief or image arrives on your task board with artifacts and full step history. You review it, then send it yourself.
Specialists ship with Jobshout. Each owns a lane, brings its own tools, and answers in the same chat behind the same approval gate. Agents are modular — a new specialist registers its prompt, tools and launch form without touching the platform.
Watch the senders you care about, research the thread, and leave a reply draft. Nothing is sent until you approve.
Evaluate a job posting against your profile and return a clear fit assessment.
Return cited findings you can open — SLAs, competitor notes, source-checked briefs.
File a draft article on the board with structure and artifacts attached.
Scan an authorised target and hand back a report with reproduction steps.
Review a GitHub pull request and leave structured feedback for your team.
Make one picture and store it on the board — no unattended posting.
Crawl a URL for title, meta, headings, robots and sitemap health. Improve proposes fixes; Publish applies via WordPress, OpsAPI or a git PR — with approval.
Every run becomes a task you can see, sort and reassign, with its output and artifacts attached. Group them into projects and plan them into sprints.
Chain specialists into a dependency graph. Steps with no prerequisites run at the same time; dependent steps wait for their inputs.
Put a job on a cron schedule — the Monday chase, the morning research brief, the weekly article. Results land on the board as usual.
Teach an agent your own tools. Jobshout speaks the Model Context Protocol, so any MCP server you run becomes discoverable tools.
OpenAI, Anthropic Claude, or a model running locally through Ollama — no API key leaving the building when you self-host.
Per-run tokens, latency and USD cost. Export OpenTelemetry traces to Langfuse when you want to go deeper.
A Go API, a Next.js interface, PostgreSQL and MinIO for object storage. Start the whole stack with docker compose up, or deploy the Kubernetes manifests. Your data and your prompts stay where you put them.
Visit jobshout.co.ukAn AI teammate you can trust to get work done. Describe the job — a sentence is enough. Explore Jobshout on the product site, or talk to Workstation about deploying the framework for your team.
Open jobshout.co.ukContact Workstation