Opportunities › Lightweight kanban for AI agents and developers › Proposed app
Tidepool Board
A small shared task board that people and AI agents both update
- Score
- 54
- confidence 63
- Evidence
- 11
- items on the page
- Rivals
- 0
- named in sources
- First version
- 25 d
- estimate
- Type
- Desktop and mobile
- PWA: on computers and phones alike
- Market since
- 2025
- New market · Verified
ScreensConcept
Desktop and mobile: Teams set up permissions and review queues on desktop, while developers and agents claim tasks and comment on cards from anywhere; both creation and approval are essential. Sample screens on a computer and a phone, in the proposed brand. Their numbers are sample data.
Sprint board
2 agents active3 agents
About this appAI proposal
Tidepool Board is a task board for developers who work with AI coding agents. Agents create, claim, move and comment on cards through MCP, and you watch the board change as they work. Every change is labelled, so you can tell who did what.
It is built to be light and local-first: run it on your own machine or a small server, and install it as a PWA that keeps working offline. A hosted option is there for teams that prefer not to run it themselves.
General kanban tools reach agents through add-on integrations. Here agents are first-class members: you set what each may do, review proposed changes and undo a run of edits in one step.
- Tagline
- One small board for you and your agents.
- Voice
- CalmPlainPrecise
- Why this name
- A tidepool is a small, contained place where many creatures share space and you can see everything at a glance, like a light board shared by people and agents.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
MCP server built in
Agents connect to a standard MCP endpoint and can list, create, claim, move and comment on cards without custom glue code.
No member yet
Agent and human attribution
Every card change shows whether a person or a named agent made it, with a full activity trail.
No member yet
Per-agent permissions
Give each agent a scope: read only, comment, move cards, or create. Revoke a key at any time.
No member yet
Review queue
Let agents propose changes that wait for your approval, or let trusted agents act directly.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#12403CDeep Kelp · primary#F3F7F4Sea Foam · surface#14201FInk · text#E2603FCoral · accent#7FC4B5Shallow · data
Typography
Aa
Tidepool Board
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
One small board for you and your agents.
PricingAI proposal
The material gives no competitor prices. These tiers are a starting point to test, based on the monetization inference of hosted, support and services offerings.
Local
Free
- Self-hosted, one board set
- MCP server
- Up to 3 agent keys
- Markdown and JSON export
Team
$12 a month
- Hosted option
- Unlimited agent keys
- Per-agent permissions
- Review queue and undo
- Activity history
Scale
Contact us
- Priority support
- Custom agent integrations
- Single sign-on
- Dedicated hosting
Would you pay Team · $12 a month?
Your answer tells PWAS which prices people accept. It changes no score.
No answers yet.
What it could be worthEstimate
What a software business making this much sells for today, from real deals. Pick a monthly revenue.
If it makes
Worth about
$194,000
Sale range $132,000 to $263,000
$120,000 a year in revenue
From 87 software businesses listed and sold on Empire Flippers in the last year, revenue vetted by the broker: they sell for about 19× their monthly revenue, most between 13× and 26×.
Measured on profit, the middle deal is 32× monthly net profit: about 2.7 years of profit.
An estimate, not an offer or financial advice. A real price depends on profit, growth, churn, how much of the work is the owner's, and the buyer.
How it would growAI proposal
How the product would reach new customers through its own use, and how the first ones would be found. A plan to test, not a result.
- Built-in loop
- Users invite others. A developer invites teammates and connects agents to a shared board. Teammates join to approve or comment on agent work, and each board is a reason to bring in more people.
- Do they become customers?
- Invited teammates are developers on the same team and can become paying seats. Test whether invited people create their own boards or only view one.
- First customers
- Solo developers and teams of 2 to 10 who already run AI coding agents such as Claude Code on daily work and want a shared task list for them.
- Where to reach them
- GitHub and open source: Publish the self-hosted core and an MCP setup guide, and answer issues in related agent projects.
- Developer communities: Share demos of agents working a board in MCP and agent-tooling forums and Discord servers.
- First milestone
- Ten teams run an agent on a shared board for two weeks, and five of them pay for the hosted Team tier.
Why people already pay for this
- VerifiedGitHub repository gained 1030 stars in 11 days
- Public claimThe tool is extremely resource-efficient, consuming only 4.5 MB of RAM and starting in 2 milliseconds, making it suitable for running on constrained environments like Raspberry Pi.
- Public claimThe tool is designed specifically for agent-human interaction, supporting both autonomous agents and human users managing tasks collaboratively.
- Public claimIntegration with Model Context Protocol (MCP) and agent frameworks suggests demand for tools that fit into agentic AI workflows.
Tech stack
Every proposal builds on the same PWA base, installable from the browser on any phone or computer. Dashed boxes are what this app adds.
- AppNext.js and TypeScript
- PWAManifest and service worker
- UITailwind CSS
- Data and authSupabase Postgres
- PaymentsStripe
- EmailResend
- HostingVercel
- This app · Agent protocolMCP server over HTTP and stdioAgents need a standard way to read and change cards with scoped keys.
- This app · RealtimeServer-sent events or Supabase RealtimeBoards must show agent changes live.
RisksAI inference
- CommercialThe core is small and the material shows an open GitHub project with fast star growth. People may prefer a free self-hosted tool and not pay for a hosted version.
- PlatformMCP is still changing. Protocol changes could break agent connections and need ongoing upkeep.
- TechnicalTwo agents editing the same card at once need claim locks and conflict handling, plus sync between local and hosted copies.
- AIA misbehaving agent could flood the board or make wrong edits. Rate limits, scopes and undo must work well from the start.
- DataLow-resource claims (4.5 MB RAM, 2 ms startup) come from another project's public claim. This product must measure its own footprint before promising it.
Everything marked AI proposal or AI inference was written by a model from the evidence on the opportunity page. The proposed name must be checked against trademark registers before it is used.