Guide
Dust vs a team skill pod: which fits your team
Dust and a team skill pod both help a team use AI the same way, from opposite ends. Dust is an agent platform: you build agents inside Dust with their own instructions, skills, knowledge and tools, connect company data from Slack, Notion, Drive and other sources (or any tool through MCP), pick among 20+ models, and people work with those agents in Dust, in Slack or through its Chrome extension. It's priced per seat with usage credits, and its Enterprise plan adds SCIM and audit logs. A team skill pod, such as AI Skill Pod, leaves people in the ChatGPT, Claude or coding agent they already use and shares the team's skills into each of them through one link. Choose Dust when you want agents that act across company data; choose a pod when you want everyone's existing AI to follow the same methods.
Short answer
- Dust: agents in Dust, over company data, many models.
- A team pod: the team's skills in the AI people already use.
- Agents on company data: Dust.
- Same methods in ChatGPT and Claude: a pod.
What each does
| Dust | AI Skill Pod | |
|---|---|---|
| What it is | A platform for building and running agents | A private skill library for a team |
| Where people work | In Dust, its Chrome extension and Slack; other AI clients can call Dust through its MCP server | In their own Claude, ChatGPT (Pro, Business or Enterprise), Claude Code, Codex, Cursor or OpenCode, per the pod's own pages |
| Company data | Connectors to Slack, Notion, GitHub, Drive and more; any tool via MCP. On Business: up to 3 connectors, 5 remote MCP servers, 5 Spaces | Not a data platform; each person's AI keeps its own connectors |
| Models | 20+ models, chosen per agent | Whatever each person's AI uses |
| Price | Per seat with credits; see Dust's pricing page, read 5 November 2026 | Free up to 3 members and 10 skills; Team $5 per seat a month from 3 seats |
Dust from its home and pricing pages, read 5 November 2026. AI Skill Pod is this site's product; its rows describe what it says it does.
A worked example: a 15-person consultancy
Illustration: a made-up business. A 15-person consultancy wants every consultant's proposals and reports written the same way. Its consultants use ChatGPT and Claude, and three associates are freelancers on their own accounts. Criteria and weights first, then the must-have, pass or fail: client material stays private. Both pass on their own pages (Dust lists SOC 2 Type II and permission spaces; the pod lists private skills), still to be checked in the trial.
| Criterion (weight) | Dust | Team pod |
|---|---|---|
| Works in the AI people already use (35) | 3: work moves into Dust, Slack and the Chrome extension; other AI clients can call Dust through its MCP server (vendor docs, unverified) | 4: Claude, and ChatGPT on Pro, Business or Enterprise (vendor claim, unverified) |
| Agents acting on company data (25) | 4: connectors and MCP, but Business caps connectors at 3 and remote MCP servers at 5 (vendor pages, unverified) | 1: not what it does |
| Cost over 2 years for 15 (20) | [check: price and credits on Dust's pricing page] | [check: price on the pod's pricing page] |
| Freelancers on their own accounts (12) | [check: guest or seat terms with Dust] | 3: invite link to any email; a freelancer on ChatGPT Free or Plus would use the pod from Claude instead (vendor claim, unverified) |
| Choice of models (8) | 5: 20+ models (vendor pages) | 3: each person's own AI |
Risks. Dust: credits vary with the model, task and tools, so cost is hard to predict; Business caps connectors at 3 and Spaces at 5; agents and their work live in Dust. The pod: a newer product; its ChatGPT route needs Pro, Business or Enterprise; it doesn't act on company data.
The totals wait for the price checks, but the job decides it here: consultants should keep writing in their own ChatGPT and Claude. The consultancy trials the pod with three consultants and one freelancer for two weeks. If it later wants an agent that answers questions from its Notion and Drive, that is a Dust-shaped job, and the two can sit side by side.
Choose
- Name the job: agents acting on company data, or the same methods in everyone's AI.
- Agents on Slack, Notion, Drive and other data: Dust.
- Same skills in ChatGPT, Claude and coding agents, contractors included: a pod.
- Price both for your seat count over two years, from each pricing page, including Dust's credits.
- Trial on a few people and turn unverified scores into tested ones.
Skills for this
- Operations and adminVendor comparison with weighted scoresCompares 2-5 vendors on weighted criteria you agree first: must-haves, evidence-based scores, total cost, risks, and one recommendation.
- FinanceSaaS spend reviewReviews SaaS subscriptions: owners, seats used vs paid, overlapping tools, renewal and notice deadlines, and cancel, downgrade or keep with savings.
- AI practiceTurn a prompt into a skillTurns a prompt that works into a skill the whole team can reuse, with a valid name, a when-to-use description, steps, an output template, an example, and checks.
Questions
What is Dust?
An agent platform. Teams build agents in Dust with their own instructions, skills, knowledge and tools, connect company data from Slack, Notion, GitHub, Drive and other sources, choose among 20+ models, and run agents on schedules or triggers. People work with the agents in Dust, in Slack or through its Chrome extension; its pricing is per seat with usage credits.
Is Dust an alternative to sharing skills in ChatGPT or Claude?
It's a different approach. Dust moves the work into Dust's own agents. A skill pod leaves people in the ChatGPT or Claude they already use and shares the same skills into them. Choose by where you want people to work and whether you need agents acting on company data.
When does a team skill pod fit better than Dust?
When people should keep their own ChatGPT, Claude or coding agent and the goal is the same methods everywhere, including contractors on their own accounts. That is the pod's own description, so trial it on a few people before relying on it.
AI skill
Have your AI compare vendors on what matters to you
Put this skill in AI Skill Pod and everyone on your team gets it in their Claude, Cursor or ChatGPT (Pro, Business, Enterprise): one invite link, free to start. Vendor comparison with weighted scores has your AI compare tools the way a buyer should: criteria and weights agreed before scoring, must-haves as pass or fail, scores with their evidence, whole-life cost over the term, risks per option, and claims only the vendor makes marked unverified.
14 days free · No card needed
Sources
- Dust — "Dust connects your company knowledge, tools, and teams so you can create, share, and run agents"; 20+ models from OpenAI, Anthropic, Google, Mistral and more; a Chrome extension (read 2026-11-05)
- Dust: Pricing — Business for teams up to 100 people, with a free tier and per-seat plans that include credits; custom agents with skills, knowledge and tools; connectors to Slack, Notion, GitHub, Drive and 20 more, or any tool via MCP; SSO; US and EU data residency. On Business, connectors reach up to 3 data sources, MCP servers 5 remote and Spaces 5; Enterprise adds unlimited connectors and MCP servers, SCIM, audit logs and single-tenant deployment. A credit is Dust's unit of AI usage, and use depends on the model, the task and the tools (read 2026-11-05)
- Dust docs: Dust in Slack — call Dust agents from Slack channels, threads and direct messages (read 2026-11-05)
- Dust docs: Dust MCP Server — MCP clients such as IDEs and AI assistants can call Dust to search workspace knowledge and work with conversations (read 2026-11-05)
- Agent Skills specification — the open SKILL.md format a pod's skills use (read 2026-11-05)