spoki-automation

Overview of spoki-automation skill

What spoki-automation is for

spoki-automation is a Claude skill for running Spoki operations through Composio’s Rube MCP toolkit. It is designed for users who want an agent to discover the current Spoki tool schemas, confirm the Spoki connection, and execute supported workflow automation tasks without hard-coding outdated API assumptions.

The key value is not a large prompt library; it is the workflow discipline: search Rube tools first, verify the spoki connection, inspect the returned schemas, then run the selected action with validated inputs.

Best-fit users and workflows

This spoki-automation skill fits teams already using Claude with MCP and Composio/Rube, especially when Spoki is part of a messaging, customer operations, or campaign workflow. It is useful when you know the business outcome you want but need the agent to identify the correct Spoki tool and current input format.

Good use cases include asking an agent to find available Spoki actions, check whether authentication is active, prepare tool calls from structured customer or campaign data, and explain what fields are required before execution.

What makes it different from a normal prompt

A generic prompt may guess tool names or schemas. spoki-automation explicitly tells the agent to call RUBE_SEARCH_TOOLS first and use the returned tool slugs, schemas, execution plans, and pitfalls. That matters because MCP tool surfaces can change, and stale examples are a common source of failed automation.

The skill is intentionally narrow: it is a connection-aware Rube MCP workflow for Spoki, not a standalone Spoki SDK, CLI, or full campaign management system.

How to Use spoki-automation skill

spoki-automation install context

Install the skill from the Composio skills repository:

npx skills add ComposioHQ/awesome-claude-skills --skill spoki-automation

You also need Rube MCP configured in your Claude-compatible client. Add https://rube.app/mcp as an MCP server, then confirm the MCP exposes RUBE_SEARCH_TOOLS. The skill depends on Rube, so installing the markdown alone is not enough.

Before running real work, ask the agent to call RUBE_MANAGE_CONNECTIONS for toolkit spoki. If the connection is not ACTIVE, complete the returned authentication flow and recheck status.

Inputs the skill needs from you

For reliable spoki-automation usage, give the agent a concrete task, the intended Spoki object or workflow, the data source, constraints, and whether it may execute actions or should only prepare a plan.

Weak prompt:

Use Spoki to automate outreach.

Stronger prompt:

Use spoki-automation to discover the current Spoki tools via Rube MCP. Check my spoki connection first. I want to prepare a workflow for sending a WhatsApp-style customer update to a provided list, but do not execute until you show me the required fields, tool slug, schema, and any missing data.

This improves results because the agent can map your goal to current tools, ask for missing fields, and avoid executing before connection and schema checks.

Recommended workflow after install

Start with SKILL.md in composio-skills/spoki-automation; it is the only support file surfaced for this skill. Read it for three operational rules: Rube MCP must be available, the Spoki connection must be active, and RUBE_SEARCH_TOOLS must be called before workflows.

A practical run sequence is:

1. Ask the agent to search tools for your specific Spoki task.
2. Review returned tool slugs, schemas, and warnings.
3. Ask it to check RUBE_MANAGE_CONNECTIONS for spoki.
4. Provide required business data in a structured format.
5. Request a dry-run plan before execution if the action affects customers.

Practical prompt pattern

Use a prompt format like this:

Use the spoki-automation skill.

Goal: [specific Spoki outcome]
Execution mode: [plan only / execute after confirmation]
Data available: [fields, list source, IDs, message content, timing]
Constraints: [do not send duplicates, require approval, language, segment]
First steps: call RUBE_SEARCH_TOOLS for this exact use case, then check the spoki connection.
Output: show selected tool slug, required schema fields, missing inputs, and proposed execution steps.

This keeps the agent anchored to discovery and validation instead of guessing a Spoki API shape.

spoki-automation skill FAQ

Is spoki-automation beginner-friendly?

It is beginner-friendly if your MCP client is already configured, but it is not a one-click Spoki automation app. You should be comfortable granting tool access, completing a connection flow, and reviewing proposed tool calls before execution. Beginners should run it in “plan only” mode first.

When should I not use this skill?

Do not use spoki-automation if you need offline Spoki documentation, direct API coding help without Rube MCP, or guaranteed support for a specific action before tool discovery. The skill’s own guidance depends on live Rube schemas, so availability is determined at runtime.

Why does the skill require tool discovery first?

The repository emphasizes RUBE_SEARCH_TOOLS because Composio tool schemas can change. Searching first gives the agent current tool names, required fields, recommended plans, and known pitfalls. This is the main safeguard against failed or unsafe automation.

Does it work without a Spoki connection?

No. The skill can help check connection status, but actual Spoki operations require an active spoki connection through RUBE_MANAGE_CONNECTIONS. If authentication is incomplete, the right next step is to follow the returned auth link, not to keep prompting the model.

How to Improve spoki-automation skill

Improve spoki-automation results with better task framing

The biggest quality gain comes from turning a vague automation request into an operational brief. Include the audience, message or object type, source data, required approvals, execution timing, and success criteria. If the task may contact customers, explicitly require a confirmation step before sending.

Example improvement:

Before executing, summarize the exact records affected, the Spoki tool slug, required fields, and any defaults you will apply. Stop if any recipient identifier is missing.

Avoid common failure modes

Common blockers are inactive Spoki authentication, missing recipient or object identifiers, ambiguous execution permission, and assuming a tool schema from memory. The fix is procedural: search tools, check connection, validate required fields, then execute only after the user confirms the plan.

If the agent skips discovery, redirect it:

Do not infer the schema. Call RUBE_SEARCH_TOOLS for this Spoki use case and base the next step only on the returned schema.

Iterate after the first output

After the first plan, ask for a gap check rather than immediately approving execution. Useful follow-ups include:

• “Which required fields are still missing?”
• “What records will be affected?”
• “What could cause duplicate or unintended sends?”
• “Show the final tool call inputs before execution.”

This turns the spoki-automation guide from a simple connector into a safer workflow automation review loop.

Repository improvements worth adding

The upstream skill would be stronger with examples for common Spoki use cases, a dry-run prompt template, and troubleshooting notes for inactive Rube or Spoki connections. A small README.md with setup screenshots or client-specific MCP configuration notes would also reduce adoption friction for new users.