The project system for AI-assisted development

Write requirements. Set priorities.
Your agents handle the rest.

AI agents need context, not copy-paste. Specboard connects your docs and tasks directly to any MCP-compatible agent—so they know what to build and what matters most.

For everyone in the AI development loop

For Product People

  • Write requirements that agents can actually read
  • See what's being worked on and what's done
  • Stay in control—agents can't close epics without you
  • No more "that's not what I meant in the spec"

For Developers

  • Get specs your AI agents can query directly
  • Stop copy-pasting context into every session
  • Let agents ask "what are the requirements?" and get a real answer
  • No more translating between PM-speak and agent prompts

Whether you're a solo dev wearing both hats, or a product person handing off to AI-augmented developers—Specboard is the coordination layer that keeps everyone (and every agent) aligned.

The handoff to AI is broken

Scattered context

Requirements in Notion, tasks in Jira, specs in Google Docs. Product people write detailed specs that never reach the agent intact.

The telephone game

Product writes spec → Dev interprets → Dev prompts agent → Agent guesses. By the time context reaches the AI, half of it is gone.

Scale breaks everything

Full products have hundreds of requirement files. Too many to dump into a prompt. No structure to know which docs matter for which task.

Links break as you iterate

Plans evolve, specs update, requirements shift. The connections between sessions, plans, and requirements drift apart.

One system. Structured context. Right docs for the right task.

Requirements agents can read

Git-backed Markdown documentation. True .md files, not proprietary blocks. Version history through Git.

Priorities agents can query

Simple kanban: Ready → In Progress → Done. Drag to rank. Agents ask "what should I work on?" and get a real answer.

Structured linking

Tasks link to their relevant requirement docs. Agent picks up a task → gets the right context, not all context.

MCP that just works

Native MCP server included. Works with any MCP-compatible agent. Query docs, query tasks, update progress.

From brief to shipped, with full visibility

1
Product

Write the spec

Create requirements in the documentation editor. Markdown files, version-controlled, human-readable.

2
Product

Set priorities

Create epics on the kanban board. Drag to rank what matters most. Link epics to their requirement docs.

3
Agent

Agent gets context

Agent connects via MCP. Queries requirements directly. No copy-paste. No re-explaining.

4
Dev + Agent

Work happens

Agent breaks down epics into tasks. Implements, tests, iterates. Progress visible to everyone.

5
Product

Review & approve

Review the PR. Mark epic complete. Agent can't close the loop—humans stay in control.

Why not just use...?

What you're usingThe problemHow we're different
Notion + JiraAI can't read either. Two systems = scattered context.One integrated system with native MCP.
Markdown in repoStatic. Gets stale. No task management. No structure.Living docs + prioritized tasks, linked and queryable.
Linear / ShortcutGreat for tasks, but no MCP. Requirements live elsewhere.Docs + tasks + MCP in one tool.
ChatPRD / AI PRD toolsGenerate docs from AI, not for AI. No agent integration.Built specifically for agent consumption via MCP.
DIY (Obsidian + Issues)Glue code. No unified context. No MCP.Purpose-built, zero integration work.

AI assists. Humans decide.

Product people write specs

They define what gets built.

Product people set priorities

They decide what matters most.

Agents propose work

They create subtasks and submit PRs.

Humans approve completion

Agents can't mark epics "done."

Everyone sees the same context

No more "that's not what I meant."

Git-backed everything

Full history, portable, no lock-in.

Built on the standard, not bolted on

MCP (Model Context Protocol) is the open standard for connecting AI agents to tools and data. We didn't retrofit MCP onto an existing product—we built MCP-first from day one.

What your agent can do

  • get_epic — Read epic details and linked specs
  • get_current_work — See what's in progress
  • get_ready_epics — Find prioritized work to pick up
  • create_task — Break down work into subtasks
  • complete_task — Mark tasks as done
  • add_progress_note — Log activity for visibility

Works with any MCP-compatible agent: Claude Code, Cursor, Windsurf, and any tool that supports the protocol.

Your data, your control

Git-backed

Your docs live in a git repository you control. Full version history built in.

True Markdown

Plain .md files, not proprietary formats. Read them anywhere, edit them with any tool.

Export anytime

Your docs are just markdown. Leave whenever you want. No data hostage situation.

Request early access

We're rolling out to select teams. Join the waitlist.

Priority onboarding
Direct founder access
Shape the roadmap
Founder pricing forever

Give your AI agents a proper brief

Stop copy-pasting context. Start shipping what you actually meant.

Request Early Access