Boards
Organize delivery with columns, tickets, priorities, epics, filters, and list view.
The shared planning system
Planning that any AI can work with.
Boards and documents for your team. Scoped access for the AI tools they choose. Every update returns to the same project record—with the author attached.
Use the AI you already have · Revoke access anytime
Current Planytic interfaceSample project data
app.planytic.com
← XA Product launch
DeliveryDEV · board
Board List
Search board
XMARSL
To do
2XA-21high
Define first workspace onboarding
3 ptsXM
XA-24medium
Create the launch architecture brief
Sep 2
Add ticket
In progress
2/3XA-25urgent
Support pasted image uploads
5 ptsXM
XA-27high
Research scoped MCP access
ai
Add ticket
Review
1/2XA-26medium
Allow projects and boards to be renamed
2 ptsai
Add ticket
Fig 01 / The project record
The plan remains useful after the conversation ends.
Planytic keeps execution and context together. Teammates work in a focused project app. Agents work through a documented, permissioned interface. Neither has to reconstruct the project from a transcript.
Documents
Keep briefs, architecture, decisions, and pasted images beside the work.
AI Access
Issue scoped tokens or connect MCP clients without locking into one model.
Fig 02 / Documents
Keep the why next to the work.
Specs, decisions, and reference material live in the project—not in someone's private chat. People can edit them directly, and agents can read or update them when their scope allows.
app.planytic.com
← Back to project
AI-reference
AI access architecture
Tris · updated 8 minutes ago
Planytic exposes project context through two controlled entry points: the AI manifest for direct API use, and the MCP server for persistent tools inside supported clients.
Request boundary
Every request resolves the token, workspace, project binding, and declared scopes before the resource endpoint is called.
Authorization: Bearer pl_agent_•••••••• GET /api/ai/whoami
Write attribution
Ticket, document, comment, and worklog mutations record the acting agent so human reviewers can distinguish automated work from teammate activity.
Fig 03 / AI Access
Connect the tool. Control the boundary.
Give any capable AI one manifest URL or connect it through MCP. The token determines which projects it can see and which actions it can take. The model stays your choice.
app.planytic.com
machine credentials
AI Access
Issue scoped tokens so AI agents (IDE, CLI, chat) can read and work your boards through the API.
How AI agents connect
Paste this URL into any AI—it documents every endpoint, convention, and rule. The agent authenticates with a token sent as an Authorization header.
https://api.planytic.com/api/ai/useme Connect via MCP
Connect your AI IDE once. It gets Planytic tools with the same scopes, bindings, and revocation as its token.
01Issue a scoped token
02Add the client config
03Open the repo and work
Active tokens
Research agent
projects:readtickets:readcomments:write
last used 4m ago
expires Sep 27
Fig 04 / Security
Access is explicit. Activity is attributable.
Agent access is treated as a credential—not a magical connection hidden behind a prompt.
- Bound access
- Limit a token to one workspace or project instead of exposing the whole account.
- Hashed secrets
- The token is shown once. Planytic stores a hash, not the reusable secret.
- Expiry and revocation
- Set an expiry when access is issued, or revoke it immediately from AI Access.
- Request-level policy
- Identity, scope, binding, workspace membership, and rate are checked on every call.
01
02
03
04
planytic
One project record. Every contributor.
Start with your team. Add the AI tools that fit the work. Keep the plan in one place.
Scoped tokens MCP support Agent attribution No model lock-in