parsehub-automation
by ComposioHQparsehub-automation is a Claude skill guide for running ParseHub workflows through Composio Rube MCP. It covers setup context, RUBE_SEARCH_TOOLS discovery, connection checks, and safe usage patterns for Web Scraping tasks.
This skill scores 68/100, which means it is acceptable for directory listing but should be treated as a lightweight integration guide rather than a complete Parsehub automation playbook. Directory users get enough evidence to understand when to use it and how an agent should begin safely through Rube MCP tool discovery, but install decisions are limited by the lack of concrete Parsehub examples, support files, or bundled implementation assets.
- Valid skill metadata clearly names the trigger domain: automating Parsehub operations through Composio's Parsehub toolkit via Rube MCP.
- Prerequisites and setup steps identify the required MCP server, RUBE_SEARCH_TOOLS availability, and the need for an ACTIVE Parsehub connection before execution.
- The skill gives an agent a repeatable operating pattern: search tools first for current schemas, check/manage the Parsehub connection, then execute workflows based on discovered tool slugs and schemas.
- No install command or local support files are provided; setup depends on manually adding the Rube MCP endpoint and completing a Parsehub connection through RUBE_MANAGE_CONNECTIONS.
- Workflow guidance appears mostly generic to Rube/Composio discovery and does not include concrete Parsehub task examples or field-level schemas, so agents must rely on live tool discovery for specifics.
Overview of parsehub-automation skill
What parsehub-automation does
parsehub-automation is a Claude skill for controlling ParseHub workflows through Composio’s Rube MCP server. It helps an agent discover current ParseHub tool schemas, verify the user’s ParseHub connection, and run automation steps such as managing scraping projects, starting runs, checking run status, and retrieving results when those actions are exposed by the current Composio ParseHub toolkit.
The key point: this skill is not a scraper by itself. It is an execution guide for using ParseHub through Rube MCP safely and correctly.
Best fit for ParseHub web scraping workflows
The parsehub-automation skill is best for users who already use ParseHub for Web Scraping and want an AI assistant to operate ParseHub tasks without manually clicking through the ParseHub UI. It fits workflows where you need repeatable scraping operations, project-run monitoring, or structured handoffs from a scraping request to an MCP tool call.
It is especially useful if your agent needs to avoid guessing tool names or input fields. The upstream skill explicitly instructs the agent to call RUBE_SEARCH_TOOLS first so it can use the latest available tool schemas instead of relying on stale examples.
Main adoption requirements
Before installing or relying on parsehub-automation, confirm that your client supports MCP and that Rube is configured as an MCP server. The skill requires:
- Rube MCP connected with
RUBE_SEARCH_TOOLSavailable - A ParseHub connection managed through
RUBE_MANAGE_CONNECTIONS - Connection status set to
ACTIVEbefore executing ParseHub operations - Willingness to let the agent discover tool schemas at runtime
If you need a standalone Python scraper, a browser automation script, or a ParseHub project designer, this skill is not the right layer.
How to Use parsehub-automation skill
parsehub-automation install context
A typical parsehub-automation install uses the skill from ComposioHQ/awesome-claude-skills:
npx skills add ComposioHQ/awesome-claude-skills --skill parsehub-automation
Then add Rube MCP to your AI client using the endpoint documented in the skill: https://rube.app/mcp. The repository source for this skill is under:
composio-skills/parsehub-automation/SKILL.md
Read SKILL.md first because this repository contains only the skill file, not extra scripts or reference folders. The important operational details are in the prerequisites, setup, tool discovery, and core workflow sections.
Required input for reliable usage
A good parsehub-automation usage prompt should include the real scraping objective and enough ParseHub context for the agent to choose the right tool after discovery. Include:
- What ParseHub project you want to operate, if known
- Whether the goal is to start a run, check a run, fetch data, or inspect projects
- Any required output format, such as CSV, JSON summary, or run-status report
- Constraints such as “do not start a new run until connection is confirmed”
- Whether the task is one-time execution or recurring monitoring
Weak prompt: “Use ParseHub to get the data.”
Stronger prompt: “Using parsehub-automation, discover the current ParseHub tools via Rube MCP, confirm my ParseHub connection is ACTIVE, find the project related to competitor pricing, start a new run if available, monitor until completion or timeout, and summarize where I can retrieve the extracted data.”
Workflow that the skill expects
The practical parsehub-automation guide is a three-step pattern:
- Discover tools with
RUBE_SEARCH_TOOLSfor the specific ParseHub task. - Check or activate the ParseHub connection with
RUBE_MANAGE_CONNECTIONS. - Execute the discovered ParseHub tool using the returned schema, not guessed fields.
This matters because Composio tool names, parameters, and recommended execution plans can change. The skill’s strongest safety feature is that it pushes the agent to search first, then act.
Tips that improve output quality
Tell the agent not to fabricate ParseHub project IDs, run IDs, or field names. Ask it to report the discovered tool slug and required parameters before execution when the action is destructive, costly, or starts a new scrape.
For multi-step scraping work, split the request into phases: connection check, project discovery, run execution, status polling, and result retrieval. This makes failures easier to diagnose and prevents the agent from mixing authentication problems with ParseHub project issues.
parsehub-automation skill FAQ
Is parsehub-automation for beginners?
Yes, if the beginner already has a ParseHub account and can complete an auth flow. The skill reduces guesswork around Rube MCP and Composio tool discovery, but it does not teach how to design a ParseHub scraping project. Users still need a working ParseHub project or enough knowledge to identify the project they want to run.
How is it better than an ordinary prompt?
An ordinary prompt may tell the model to “use ParseHub,” but it may guess outdated tool names or parameters. The parsehub-automation skill gives the agent a specific operating rule: search tools first, check the ParseHub connection, then execute using the current schema. That makes it more reliable for live MCP usage.
When should I not use this skill?
Do not use parsehub-automation if you need to build a custom crawler outside ParseHub, bypass ParseHub limits, scrape sites without respecting terms, or automate a browser directly. It is also a poor fit if your AI client cannot connect to MCP servers or if your organization does not allow third-party tool connections through Composio/Rube.
Does it include scripts or templates?
No. Repository evidence shows a single SKILL.md file and no bundled scripts, rules, references, or assets. That is acceptable for a tool-orchestration skill because the main value is the runtime workflow, but users looking for ready-made ParseHub project templates will not find them here.
How to Improve parsehub-automation skill
Improve parsehub-automation prompts with task specifics
The biggest improvement is to provide concrete task context. Instead of asking for “ParseHub automation,” specify the desired action and success condition:
“Discover ParseHub tools, verify connection, list available projects, identify the project named Retail Price Monitor, start a run only after confirming the project match, then return run ID, status, and next check time.”
This gives the agent decision points and reduces the chance of operating the wrong project.
Guard against common failure modes
Common failures include inactive ParseHub connections, missing project identifiers, stale assumptions about tool parameters, and long-running scrape jobs. Ask the agent to expose each phase:
- “Show the connection status before continuing.”
- “Use
RUBE_SEARCH_TOOLSbefore selecting a tool.” - “If a required field is missing, ask me instead of guessing.”
- “If the run is still processing, report status and recommended retry interval.”
These instructions align directly with how the skill is designed.
Iterate after the first output
After the first run, improve the workflow by capturing stable identifiers, preferred output formats, and monitoring expectations. For example, if the agent successfully finds a ParseHub project ID, reuse it in future prompts. If the output is too verbose, request a compact run report with only project name, run ID, status, started time, and result link if available.
Add local team conventions
Teams can make parsehub-automation more dependable by documenting internal ParseHub project names, ownership, allowed run windows, and data retention rules. Put those details in your prompt or a local instruction file so the agent knows which projects are safe to run and when human approval is required.