
Agentistics MCP Server: Usage Stats and Task Board for Coding Agents
Lets Claude Code and other assistants read your AI coding usage and cost, and run a shared task board and session groups.
This is the MCP side of agentistics, an app that reads the session files your coding assistants write on your computer. It does not work alone. The tools forward requests to a running agentistics server, which is on port 47291 by default. The agentistics installer sets that server up and adds this MCP server to Claude Code for you.
Beyond usage numbers, it can create and edit tasks, send messages between sessions, and move a session to another account. Those tools are meant for agents that manage other agents, so read the safety section before you give an assistant all 48.
Good for
- People who use several coding assistants and want to ask which one costs most
- Developers running a fleet of agent sessions that need a shared task board
- Teams that already run the agentistics dashboard and want it inside the assistant
Works in
12 should work
Tap an app to see how to set it up and any known problems.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Supports local MCP servers.
Client does not support this transport.
Client does not support this transport.
Install Agentistics MCP
Claude Code
- 1
Copy this prompt.
Add the Agentistics MCP server to Claude Code. Run this command: curl -fsSL https://agentop.openvibes.tech/cli | sh Then check that it loaded and tell me if I need to sign in or restart.
- 2
Paste it into Claude Code. It shows the change and asks before running it.
Requires Linux, macOS or WSL (README).
Run the agentistics installer first. The agentop server it starts adds this server to Claude Code for you at user scope, so the first method needs no extra step. The npx methods only work while that server is running, on port 47291 unless you set AGENTISTICS_API.
From the project README. Source.
Tools
Shows your total AI coding usage and cost
Returns all-time totals across the coding assistants agentistics tracks: tokens, estimated cost, sessions, streak, most used model and top project. You can narrow it to one assistant.
Try: “How many tokens and dollars have I spent on coding assistants overall?”
Read ·
agentistics_summaryCompares your coding assistants side by side
Lists each tracked assistant (Claude Code, Codex, Gemini, Copilot and others) with its sessions, messages, tokens, estimated cost and last active date.
Try: “Which of my coding assistants costs me the most?”
Read ·
agentistics_harnessesBreaks cost down by model
Gives token counts (input, output, cache read and write) and an estimated dollar cost for each model, for all assistants or one.
Try: “Show cost per model for Claude Code only”
Read ·
agentistics_costsLists your projects with usage and cost
Returns each project with its session count, messages, tokens, estimated cost and last active date. You can scope it to one assistant.
Try: “Which project used the most tokens this month?”
Read ·
agentistics_projectsLists your git repos with usage and cost
Groups usage by normalised git remote, so the same repo on different machines or paths is counted once. Sessions with no linked repo go under "unlinked".
Try: “Which repositories cost me the most?”
Read ·
agentistics_reposShows your most recent sessions
Returns the latest sessions with assistant, project path, duration, message count, tokens and model. First prompts in the result are the user's own text, so treat them as untrusted input.
Try: “List my last 10 sessions and their cost”
Read ·
agentistics_sessionsChecks your subscription limits before starting work
Returns subscription budgets, pacing and a routing recommendation for which assistant and model to use next. Its description tells the agent to read it before opening sessions and not to start new workers on a plan that is on hold.
Try: “Which assistant has room on its plan right now?”
Read ·
agentistics_plan_limitsLists your saved tags with totals
Lists tags (saved filters over repos, projects, machines, teams or accounts) with sessions, cost and tokens added up for each.
Try: “What tags do I have and what did each cost?”
Read ·
agentistics_tagsOpens one tag in detail
Returns one tag by name or id with totals, breakdowns by project, model, harness, repo and member, a daily activity series and the date window.
Try: “Break down the "client-work" tag by model”
Read ·
agentistics_tag_detailLists the widgets you can add to a dashboard layout
Returns a fixed list of dashboard components with ids, categories, descriptions and default sizes. It calls nothing else. Use it before building or editing a layout.
Try: “What chart widgets can I put on my custom dashboard?”
Read ·
agentistics_component_catalogShows your saved dashboard layouts
Returns each custom layout with its components, which one is active, and the instance id of each component so one can be removed.
Try: “Show my dashboard layouts”
Read ·
agentistics_get_layoutsMakes a new empty dashboard layout
Adds a named, empty layout to the custom dashboard page.
Try: “Create a layout called "Weekly review"”
Write ·
agentistics_create_layoutBuilds a whole dashboard in one step
Creates a layout, adds all the components you list with automatic positions, and can switch to it right away.
Try: “Build a cost dashboard with the summary, per-model costs and top projects, and make it active”
Write ·
agentistics_build_layoutAdds one widget to a layout
Puts a single component on a layout. It is placed at the bottom of the grid unless you give x and y positions.
Try: “Add the costs widget to my "Weekly review" layout”
Write ·
agentistics_add_componentRemoves one widget from a layout
Removes a component from a layout by the instance id that the layouts list shows. The layout settings are saved in agentistics preferences.
Try: “Remove the streak widget from my weekly layout”
Write ·
agentistics_remove_componentSwitches which layout the dashboard shows
Makes a named layout the active one on the custom dashboard page.
Try: “Switch my dashboard to the "Weekly review" layout”
Write ·
agentistics_set_active_layoutDeletes a saved dashboard layout
Deletes one named layout. The last remaining layout cannot be deleted. No usage data goes with it, only the saved arrangement of widgets.
Try: “Delete the old "Test" layout”
Delete ·
agentistics_delete_layoutGives you a link to a PDF usage report
Builds a local dashboard link that opens the PDF export screen for a date range. It exports nothing itself; you review the settings and click Download.
Try: “Give me a link to a PDF report for the last 30 days”
Read ·
agentistics_export_pdfLists every task on the agent task board
BETA. Returns all tasks on the board with cost, rounds, sessions and tokens used, plus a board-wide overview. The description says to start here to see what work exists.
Try: “What tasks are on the board and what has each cost so far?”
Read ·
agentistics_tasksOpens one task in full
BETA. Returns a task with its attempts, sessions filed under it, comments, subtasks, files, links and metrics. You can give the task id or its exact title.
Try: “Open the "Fix login bug" task and show its comments”
Read ·
agentistics_taskCreates a task
BETA. Adds a task with a title and optional description. If a task with that title already exists it returns that one instead of making a duplicate.
Try: “Create a task called "Add dark mode" with a short description”
Write ·
agentistics_task_createChanges a task's fields
BETA. Sets title, detail, priority (urgent, high, medium or low), type, dates and labels. Fields you leave out stay as they are, and an empty string clears one.
Try: “Set the priority of "Add dark mode" to high and add a due date”
Write ·
agentistics_task_editMoves a task to another status
BETA. Changes a task's status, for example to in progress or done. Moving a task to blocked needs a reason or a list of blockers.
Try: “Mark "Add dark mode" as done”
Write ·
agentistics_task_statusLists the board's statuses
BETA. Returns every status on the board with its label, colour and whether it is protected. A fresh board has todo, in_progress, blocked and done.
Try: “What statuses does my board have?”
Read ·
agentistics_task_statusesAdds, renames or removes a status
BETA. Creates a status, changes a label or colour, or removes a status. The server refuses to remove the four protected statuses or any status still in use.
Try: “Add an "in review" status”
Delete ·
agentistics_task_status_editLists the board's task types
BETA. Returns the task type labels, colours, order and how many tasks use each.
Try: “What task types exist?”
Read ·
agentistics_task_typesAdds, renames or removes a task type
BETA. Creates a type from a label, edits a label or colour, or removes a type. Removal is refused while any task still uses it.
Try: “Add a task type called "Research"”
Delete ·
agentistics_task_type_editLeaves a comment on a task or subtask
BETA. Adds a note to a task, or to one of its subtasks or groups. It can carry attachments as references to files in the chat's attachment store.
Try: “Comment on "Add dark mode" that the CSS part is finished”
Write ·
agentistics_task_commentAdds or edits a subtask
BETA. Adds a subtask by title, or edits one by id: done, status, dates, blockers or parent group. A subtask can be blocked by another subtask of the same task.
Try: “Add subtasks "Design colours" and "Update components" to "Add dark mode"”
Write ·
agentistics_task_subtaskAttaches or removes a link on a task
BETA. Adds an http or https link, such as a pull request or issue, with an optional label and kind. Passing a link id with remove detaches it.
Try: “Attach this pull request link to "Add dark mode"”
Delete ·
agentistics_task_linkSets which tasks must finish first
BETA. Replaces the list of tasks that block this one. A task cannot block itself.
Try: “Block "Release" until "Add dark mode" and "Fix login bug" are done”
Write ·
agentistics_task_blocked_byShows which tasks an agent can pick up now
BETA. Returns open, unblocked tasks that no other agent holds, most urgent first, and lists tasks that are withheld with the reason. It claims nothing.
Try: “What should I work on next on the board?”
Read ·
agentistics_task_nextReserves a task for an agent
BETA. Takes a time-limited lease on a task so two agents do not do the same work. It expires after 30 minutes by default, can be refreshed, and can be released when done.
Try: “Claim "Add dark mode" so no other agent starts it”
Write ·
agentistics_task_claimShows recent task board activity
BETA. Lists status moves, claims, releases, priority changes and filed sessions, newest first, for one task or the whole board.
Try: “What happened on the board in the last day?”
Read ·
agentistics_task_activityFiles a session under a task
BETA. Links a session to a task or one of its subtasks so the task's cost, tokens and rounds come from it. It can also detach a session.
Try: “File this session under "Add dark mode"”
Write ·
agentistics_task_sessionDeletes a task and its history
BETA. Deletes a task together with its comments, subtasks, files and links. The sessions filed under it are kept. The project's own policy table rates this as a destructive action.
Try: “Delete the task "Old experiment"”
Delete ·
agentistics_task_deleteLists your session groups
Lists the named folders in the Sessions sidebar with the sessions in each. It works on a machine, not on a central server.
Try: “Which session groups do I have?”
Read ·
agentistics_session_groupsCreates a session group
Creates a named group and can file sessions into it in the same call. If any session reference matches nothing or is ambiguous, nothing is created.
Try: “Create a group "Refactor" with my two refactor sessions”
Write ·
agentistics_session_group_createAdds, removes, renames or deletes a group
Files a session into a group, takes it out, renames a group, or deletes a group. Deleting a group only ungroups its sessions; the sessions are not deleted.
Try: “Move this session into the "Refactor" group”
Delete ·
agentistics_session_group_editTurns a session's notifications on or off
Reads or sets the notification switch for one session. Muting removes the bell, sound and toast, but the session still shows as waiting and its events are still recorded.
Try: “Mute notifications for my long-running test session”
Write ·
agentistics_session_notifySends a message to another session
Types a message into another agentistics session, tagged as a handback, a block or a question. If the sending session is filed on a task, the message is also saved there as a comment.
Try: “Tell the parent session that the migration is committed”
Write ·
agentistics_session_messageGets or sets a session's role
Reads or sets a session as leader, worker or auto. Setting leader renames the session with a "[ LEADER ]" prefix. A session can change only itself or sessions it started.
Try: “Make this session the leader”
Write ·
agentistics_session_roleContinues a session on another account
Lists accounts a session can move to, or copies its conversation to another account and resumes it there. Useful after a usage limit. Resuming the same conversation works for Claude Code and Codex only.
Try: “My session hit its limit, move it to my other account”
Write ·
agentistics_session_switch_accountReads or sets away mode
Reports or switches away mode, which tells newly started sessions the owner is out and what the owner pre-approved (releases, workers, merges). Only a verified leader session can write it.
Try: “Turn on away mode until tomorrow morning”
Write ·
agentistics_session_awayRecords a leader check-in report
Saves a plain-language progress report with a headline, delivered and in-progress items, next steps and health rows, and adds a check-in event to the filed task.
Try: “File the check-in report for this round”
Write ·
agentistics_checkin_reportRegisters a durable wait (experimental)
Experimental, tied to the AGENTISTICS_FLOWS setting. Registers a wait for a handback, free memory, a time or a gate, and the server types a wake message into the session when it is met. Can also acknowledge, cancel or list waits.
Try: “Wake me when the worker session hands back”
Write ·
agentistics_waitRuns a built-in read-only flow (experimental)
Runs one of the official built-in flows. The description calls flows experimental and read-only. It is hidden unless the server reports flows as on, and refuses when AGENTISTICS_FLOWS is off.
Try: “Run the flow that closes delivered workers”
Read ·
agentistics_flow_runChecks a flow run (experimental)
Reads the status of a flow run by its run id. Hidden unless flows are on.
Try: “What is the status of flow run abc123?”
Read ·
agentistics_flow_status
Read from the source code
Permissions
high risk · 5 things to watch
High risk
The assistant can delete tasks, write into your other agent sessions and change dashboard settings in your agentistics server. Use it with a trusted local server and decline delete and session-message calls you did not ask for.
What it can do
- Read your data
- Create and change things
- Delete things
- Read-only mode
Watch out for
It can delete tasks and their history
agentistics_task_delete removes a task with its comments, subtasks, files and links. Sessions filed under it are kept. Decline that call unless you meant it. unpkg.com
Agents can type into your other sessions
agentistics_session_message writes into another session's conversation, and agentistics_session_switch_account copies a conversation to another account. Review those calls before you allow them. unpkg.com
Away mode can pre-approve releases and merges
agentistics_session_away records what the owner pre-approved while away, such as releases, workers and merges, and tells new sessions about it. The server accepts writes only from a verified leader session. unpkg.com
Every tool call is logged to a file
The server appends a record of each call (tool name, risk class, targets, a hash of the arguments, client and outcome) to agent-audit.jsonl in ~/.agentistics, or in AGENTISTICS_DIR if set. The file is created with owner-only permissions. unpkg.com
Task text and prompts are untrusted input
Tool descriptions say first prompts and task comments may be untrusted text. Another session or person can write instructions there that your assistant then reads. unpkg.com
Security advisories
No published security advisories as of Nov 1, 2026.
Technical details
| Turning on read-only | No read-only switch is documented or in the package. 21 tools only read, but every tool is advertised. |
|---|---|
| Talks to | Only the agentistics server you point it at, http://localhost:47291 by default |
| Sign-in and keys | No key or sign-in. The tools trust whatever agentistics server answers at the address in AGENTISTICS_API. |
Questions
Is the Agentistics MCP server official?
Yes. It is published from the agentistics project's own repository and is the MCP side of the agentistics app.
Does it work without the agentistics server?
No. It forwards every request to a running agentistics server, on port 47291 by default. Start the server first, or set AGENTISTICS_API to its address.
Which assistants does it work with?
Claude Code is the main target, and the agentistics server registers it there for you. The package README also gives a Claude Desktop config. The agentistics docs say the server adds the MCP to Codex, Gemini and Copilot CLIs too.
Does it need an API key?
No. It sends no credentials. It only needs the address of your agentistics server.
Can I run it read-only?
No. No read-only setting exists in the package. The tools that change things are the task, session group, layout and session tools.
Does it send my data anywhere?
The MCP server talks only to the agentistics server address you give it. The agentistics README says that on a single machine the app opens no outbound connection, and in Team Mode a machine sends computed metrics only.
Why does the package README say 13 tools?
It is out of date. The published package registers 48 tools; two of them (the flow tools) stay hidden unless the server reports flows as on.
What is the license?
FSL-1.1-ALv2 according to the package metadata, which is source-available and not open source. The npm package does not ship a license file.
Information
updated Nov 2, 2026
- Publisher
- Official
- Package
@agentistics/mcp- Registry
- npm
- Transport
- stdio
- Auth
- None
- Version
- 2.173.1
- License
- FSL-1.1-ALv2
- Last commit
- Oct 31, 2026
- Open issues
- 65
Sources
Docs, issue trackers and discussions this page is built from. Issues, quotes and findings above link to their own source.
Show all 4Hide
- @agentistics/mcp README · github.com
- agentistics README · github.com
- agentistics MCP docs · github.com
- @agentistics/mcp 2.190.0 dist · unpkg.com