Building Automations logo

Building Automations

Organization
rejot-dev
building-automations

Automate persistent Backoffice behavior from events or schedules. Use when the user explicitly asks to create or change an automation, event-triggered route, scheduled route, or saved workflow that should run later.

Overview

Publisherrejot-dev
Repositoryfragno
Skill namebuilding-automations
Stars
62
Forks
6
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by rejot-dev on GitHub. Read the source before you install it.

Installation

Install the Building Automations AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/rejot-dev/fragno.git /tmp/fragno
mkdir -p .claude/skills
cp -r /tmp/fragno/apps/backoffice/content/static/skills/building-automations .claude/skills/building-automations
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Building Automations in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Building Automations on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Building Automations is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Building Automations

Treat every automation as a route with one trigger → action. Event triggers select catalog events; scheduled triggers own their cadence. The action starts or signals a saved workflow or forwards the event.

Required process

  1. Read the live declarations before editing:

    • "/static/codemode/providers/events.d.ts" for the event catalog;
    • "/static/codemode/providers/router.d.ts" for route schemas and actions;
    • "/static/codemode/workflow-authoring.d.ts" when saving a workflow;
    • "/static/codemode/providers/store.d.ts" only when durable key/value coordination is needed.

    Then inspect the event descriptor and existing artifacts:

    js
    async () => {
      const catalog = await events.catalogList({});
      const eventDescriptor = await events.catalogGet({
        source: "telegram",
        eventType: "message.received",
      });
      const routes = await router.list({});
      const workflowFiles = await state.glob({
        pattern: "/workspace/automations/**/*.workflow.js",
      });
      return { catalog, eventDescriptor, routes, workflowFiles };
    };

    Complete when the exact catalog source and eventType, relevant existing routes and workflow files are known.

  2. Choose the smallest topology that satisfies the request:

    • External or product event: route with an event trigger.
    • Time-based trigger: route with a schedule trigger.
    • New work: start_workflow.
    • Resume waiting work: send_workflow_event.
    • Cross-scope delivery: forward_event.

    Use stable, namespaced ids. For start_workflow, make workflowScriptPath point to a saved file containing exactly one statically named defineWorkflow(...). Backoffice validates the authored name from the file and keeps workflow-host routing internal. Complete when every trigger, action, file path, and instance-id template has one unambiguous value.

  3. Check prerequisites. When an external capability is involved, inspect it with connections.get({ id }). If configuration is incomplete, use the Configuring Connections skill and collect the missing values from the user. Complete when every provider the automation will call is configured or the user has been told exactly what remains missing.

  4. Create the required artifacts in dependency order: saved workflow, then route.

    Save workflow implementations in /workspace/automations/ with a .workflow.js suffix:

    js
    async () => {
      const path = "/workspace/automations/telegram-hello.workflow.js";
      await state.writeFile({
        path,
        content: `defineWorkflow(
    { name: "telegram-hello" },
    async (event, step) => {
     return await step.do("capture-event", async () => {
       return { receivedEventId: event.id };
     });
    },
    );
    `,
      });
      return { path };
    };

    Grants

    Choose authority from the workflow's user source:

    • delegated-user: the triggering event already carries an internal user principal.
    • linked-user: resolve the external initiator's active user binding.

    For either user-backed mode, use grants: "inherit" when the workflow should retain the user's current permissions. The route remains a delegate, while role changes, bans, membership removal, and route disablement still apply. Use an explicit canonical grant array only when the route must narrow the user's authority.

    organization-automation has no user principal to inherit from. Give it the explicit canonical permissions used by its runtime tools. Creating or updating an explicit grant array requires the configuring execution to hold every requested grant. Complete when the authority mode names the intended principal and its grant policy is either inherited user authority or the smallest explicit automation grant set.

    Signal an existing durable workflow on a schedule, with an explicit IANA time zone:

    js
    async () =>
      await router.create({
        id: "daily-digest",
        name: "Daily digest",
        trigger: {
          kind: "schedule",
          cadence: {
            kind: "cron",
            expression: "0 9 * * 1-5",
            timeZone: "America/New_York",
          },
        },
        action: {
          kind: "send_workflow_event",
          target: { kind: "instance_id", template: "daily-digest" },
          eventType: "scheduled",
        },
      });

    Create event-triggered routes with the router provider:

    js
    async () => {
      return await router.create({
        id: "telegram-hello",
        name: "Telegram hello",
        trigger: {
          kind: "event",
          source: "telegram",
          eventType: "message.received",
          matcher: { path: "$.payload.text", op: "startsWith", value: "/hello" },
        },
        priority: 1000,
        action: {
          kind: "start_workflow",
          authority: { kind: "linked-user", grants: "inherit" },
          workflowScriptPath: "/workspace/automations/telegram-hello.workflow.js",
          instanceIdTemplate: "telegram-hello-${event.id}",
        },
      });
    };

    For send_workflow_event, identify the durable run only by its instance ID. Use { kind: "instance_id", template } when the event can render the workflow instance id directly. Use { kind: "stored_instance_id", keyTemplate } when a prior run stored the instance id under a rendered store key. Complete when every required artifact has been created or updated successfully.

  5. Use the automation store only for small durable coordination values. Values are strings, structured values use JSON.stringify, and keys are stable and namespaced. Store inputs contain domain data only; trusted execution provenance is recorded by the kernel. Categories are ordinary labels. Use verification for JSON text that must satisfy a schema.

  6. Re-read the completed route with router.get({ id }) and state.readFile({ path: workflowScriptPath }) when saved. Run connections.verify({ id }) for external providers. Complete only when the route trigger, action, workflow name, file path, and connection status all line up end to end.

Legacy router.cm.js files are outside this topology. Database-backed router rules are the routing source of truth; saved workflow files contain the behavior.

Frequently asked questions

What does the Building Automations AI skill do?

Automate persistent Backoffice behavior from events or schedules. Use when the user explicitly asks to create or change an automation, event-triggered route, scheduled route, or saved workflow that should run later.

Why use Building Automations on TypingMind?

Because you install it once and use it with any model. Building Automations is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Building Automations in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/rejot-dev/fragno/tree/main/apps/backoffice/content/static/skills/building-automations. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Building Automations?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Building Automations?

As many as you like. As long as a model supports skills, you can use Building Automations with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Building Automations AI skill free?

Yes. It is published on GitHub by rejot-dev under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇