Agent teams, swarms, memory and hooks for Claude Code and Codex, served over MCP.
Self-Starter is installed from its publisher's own source and answers where it runs, so this marketplace is not in the path of a single call. There is no address here to send one to, and a panel that pretended otherwise would be showing you an answer we made up. Install it and call it from your own client — the Installation tab has the entry for each one.
What it does
Local zero-token event triggers and standing jobs for Claude Code, Codex, Antigravity & OpenCode.
Quickstart
# 1 — add it to Claude Code
claude mcp add --scope user self-starter-mcp '--' npx -y self-starter mcp
# 2 — ask your agent for something these tools do
# trigger_draft, trigger_recipes, trigger_activate and 11 more
Self-Starter costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.
Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.
Where are you running it?
Every route below installs the same thing and ends at the same approval screen. This one runs on your machine: your client starts Self-Starter as a process under your own user, with your files and your network, so the tool surface below is what it can do to your computer rather than to a server somewhere else. Its scan result is on the Security tab — read the surface before you approve it.
This is a public server: you run it yourself and this marketplace is not in the path. You need Node.js — the current LTS (24) is the safe choice; npx comes with it. Claude Code registers it in one command; --scope user makes it available in every project. The '--' is in quotes so that PowerShell passes it on.
claude mcp add --scope user self-starter-mcp '--' npx -y self-starter mcpReconnect, or start a new session, and the tools appear in the model’s tool list.
One config entry your client uses to start the process locally. A local server runs with your file system and your network, which is why it is priced without metering.
14 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.
Tool surface
What the model actually sees. Descriptions are diffed on every release — see version history.
Read from its source code. A server that builds its tool list when it starts can list more tools than this.
Submit a new trigger, or a new version of one, as a draft. The engine checks the definition, saves the probe code and runs the probe once.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| definition | object | yes | The trigger definition: id, name, intent, probe, mode, for, policy, action, target, notify (see the skill). |
| code | string | no | The probe's TypeScript source for Deno. It reads one JSON object on stdin and prints one JSON object on stdout. Unless the definition sets probe.access to sandbox, it runs with the user's permissions: |
| recipe | string | no | Id of a ready-made recipe from trigger_recipes, such as github-ci, log-watch, http-health, futu-news or ibkr-account. The recipe supplies the code, access, permissions and mode; set its parameters in |
| replace | boolean | no | Set to true to replace the existing trigger with the same id (needed to change its code or mode). |
List the ready-made recipes: probes that ship with the plugin for common things to watch, such as GitHub PRs and CI runs, local log errors, HTTP health endpoints, Futu stock news, or an Interactive Brokers account.
No parameter schema on file.
Turn a draft on: its probe starts running on schedule, and each fire hands the task back to you. The trigger keeps working on its own after this,
| Parameter | Type | Required | Notes |
|---|---|---|---|
| draft_id | string | yes | The draft id returned by trigger_draft. |
Change an existing trigger: probe.params, probe.interval, probe.timeout, probe.access, probe.permissions, for, policy, action, target (kind, cwd),
| Parameter | Type | Required | Notes |
|---|---|---|---|
| trigger_id | yes | — | |
| changes | object | yes | The fields to change, shaped like the definition, for example {"probe": {"params": {"threshold": 210}}}. null removes a field. |
Make a trigger that fires more often than is useful quieter, keeping its state (no new baseline, no fire again for a condition that already holds):
| Parameter | Type | Required | Notes |
|---|---|---|---|
| trigger_id | yes | — | |
| reason | string | yes | Why, in one sentence the user will read, such as: fired 14 times in the last hour on the same flapping health check. |
| check_every | string | no | Check no more often than this, such as 15m. Only longer than now. |
| cooldown | string | no | After a fire, wait this long before the next one, such as 1h. Only longer than now. |
| for | string | no | State triggers: how long the condition must hold before it fires, such as 5m. Only longer than now. |
| max_fires | integer | no | Stop after this many fires in total. Only lower than now. |
| pause | boolean | no | Pause the trigger until the user resumes it. |
List all triggers with their state, last check, next check and number of fires.
No parameter schema on file.
Details of one trigger: its definition, the state of its firing rules (waiting, cooling down, ...), recent checks and recent fires.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| trigger_id | yes | — |
Run an active trigger's probe once now and show the result, without firing or changing anything.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| trigger_id | yes | — |
Pause a trigger: its probe stops running until it is resumed.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| trigger_id | yes | — |
Resume a paused trigger. Its probe runs again right away.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| trigger_id | yes | — |
Delete a trigger, its probe code and its state. Its past fires stay in the history.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| trigger_id | yes | — |
Recent fires, newest first, with how each was handled and the receipt. Leave out trigger_id to see all triggers.
| Parameter | Type | Required | Notes |
|---|---|---|---|
| trigger_id | no | — | |
| limit | number | no | How many fires to show (default 10, at most 100). |
Report that you finished handling a trigger fire (a message that starts with [self-starter]). Call it once per fire, at the end,
| Parameter | Type | Required | Notes |
|---|---|---|---|
| fire_id | string | yes | The fire id from the [self-starter] message, such as fire_... |
| summary | string | yes | One line for the user: what you found or did. |
| notify | boolean | yes | true if the user should get a notification now. |
Check the Self-Starter background service: version, available agent CLIs, trigger counts and today's agent runs.
No parameter schema on file.
- Every tool, no call limit
- No card, and mcprush charges nothing for it
- Source published under MIT
- Runs on your machine — nothing of it reaches our gateway
- Nothing to cap, because mcprush bills nothing
What counts against your monthly calls
| Tool | Unit | Calls used | Out of the allowance |
|---|
Nothing here is billable: mcprush charges nothing to install Self-Starter or to call it, at any volume.
Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.
| Term | Level | What it prices | Points |
|---|---|---|---|
| capability-exposure | high | capability blast radius (high) — client exposure if the model is manipulated | −6 |
| verification-discount | source | publisher verification (provenance) — cryptographic build provenance ties the artifact to its source | −0 |
| coverage-honesty | source | inspection depth (source) — how much of the target the scan could see | −0 |
What the scan could actually read
A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.
Tools were statically extracted from the published source (14 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".
Capability — what it could do if the model were manipulated
Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. high is the level these add up to.
| Tool | Capability tags | Why the tag was assigned |
|---|---|---|
| trigger_draft | external-sink | external-sink: declares readOnlyHint: false, openWorldHint: true |
| trigger_recipes | no tags | |
| trigger_activate | state-write | state-write: declares readOnlyHint: false |
| trigger_update | state-write | state-write: declares readOnlyHint: false |
| trigger_tune | state-write | state-write: declares readOnlyHint: false |
| trigger_list | no tags | |
| trigger_get | no tags | |
| trigger_test | untrusted-input | untrusted-input: declares openWorldHint: true (interacts with entities outside the server) |
| trigger_pause | state-write | state-write: declares readOnlyHint: false |
| trigger_resume | state-write | state-write: declares readOnlyHint: false |
| trigger_delete | state-write | state-write: declares readOnlyHint: false, destructiveHint: true |
| trigger_history | no tags | |
| trigger_report | state-write | state-write: declares readOnlyHint: false |
| service_status | no tags |
Toxic-flow graph
The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.
Supply chain and provenance
This is the first scan of this surface here, so there is nothing yet to compare it against.
Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.
Release history
Always latest — the install command names no version, so it installs the newest release when it runs; the current release on file is 0.1.1. Pin a release below and the command asks for exactly that one.
No release note was published with this version.
One review per account, from a signed-in account that does not publish this listing. Nothing else is asked — you can say what you think before you install it. Publishers can reply once per review, and a review can be edited or deleted by whoever wrote it.
Any signed-in account that does not publish Self-Starter can review it, once — before installing it or after. A review from an account that has installed it is marked as one.
Nobody has reviewed this listing. The rating on the card is the mean of the reviews written here and nothing else, so there is no rating until somebody writes the first — which takes a signed-in account that does not publish it, and nothing else.