Install
$ agentstack add skill-intuned-skills-agent ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Intuned
When this skill runs, orient the user: give them a short, friendly overview of what Intuned is and what they can ask for, then point them at the right next step based on what they want. Keep it user-facing — talk about what you'll do for them, not internal mechanics (skills, sub-agents, APIs-as-files). Don't dump this whole document back; adapt it to what they asked.
What Intuned is
Intuned turns web interactions into reliable, callable automations. The user describes what they want done on a website — extract data, fill a form, complete a multi-step workflow — and you build real Playwright code (Python or TypeScript), tests it locally, and gets it ready to run at scale on the Intuned platform.
The user doesn't need to know Playwright, selectors, or the project layout. The user points at a site, say what you want, and you do the building.
You must understand that you will write an automation that will do scrapping, not do the scrapping directly, You must also be clear with the user with that, so if the user wants to extract all the data from a website, you will write the reliable automation code to do that, then you and the user will be able run that code to get the data.
What the user can ask for
- "Automate / scrape / extract … from " — build a brand-new
automation. Ypu explores the site, proposes a plan, and build it.
- tests it once the user approves. →
create-intuned-project - "Change / add / fix something in my project" — modify an existing Intuned
automation: add an API to a project, adjust fields, handle a new page, fix a break. → edit-intuned-project
- "Test my project / run it end-to-end" — run the automation locally or as a
platform test job. → test-intuned-project
- "Something's failing / debug this" — investigate and fix a broken run. →
investigate-and-fix, and trace-debugging for a trace file (.zip).
- "How does X work in Intuned?" — concepts (projects, APIs, jobs,
attachments). → intuned-overview
Recommended patterns
- Start from a real URL and a concrete goal. "Get all job listings from
example.com/jobs, with title, company, and link" beats "scrape a job site."
- explore before building. look at the actual page and
come back with a plan (fields, pages, pagination) for the user to approve — review that plan; it's the cheapest place to course-correct.
- Parameterize instead of hardcoding. Page limits, dates, search terms, and
counts should be inputs the user can change per run — you should ask about defaults rather than baking values in.
- Prefer splitting work into focused automations (e.g. list → detail) over
one giant script — easier to test, reuse, and run in parallel.
- Keep secrets out of chat. Logins, API keys, and tokens go into the project's
parameter/env files (set up placeholders for the user to fill), never pasted into the conversation.
- Build and test locally first, deploy when you're happy. Nothing goes live
until the user chooses to deploy.
Getting started
The plugin drives the intuned CLI, so it must be installed and signed in:
npm install -g @intuned/cli
intuned auth login
If the user hits auth or "command not found" errors, point them here first.
It is also recommended to have an Intuned project setup and ready to go, intuned CLI uses a browser, it will use playwright's chromium if it is avialble. If you are working in a directory that doesn't have an intuned project initialized, then the CLI will fail to start browser since it can't find the chromium executable.
Direction
Based on the user's requests, you can go into different directions:
- Create a new automation: Use
/intuned:create-intuned-projectskill. - Edit an existing project: The user must point to the project directory to work with, Use
/intuned:edit-intuned-projectskill. - Investigate an issue or an error: The user must provide what to investigate and a project directory for the project to fix. Use
/intuned:investigate-and-fixskill. - Generic Intuned related questions: The user wants to ask something about intuned, like auth-sessions, bot detections, jobs, runs, or anything related to intuned in general.
Going deeper
For the underlying concepts — how projects, APIs, jobs, and attachments fit together — see the intuned-overview skill. For platform docs, the agent can search them via the intuned-docs tools.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: Intuned
- Source: Intuned/skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.