add-portal

Generate a job-portal search skill with a tested CLI for any local job board.

Updated Sep 23, 2026
One-click install
npx skills add https://github.com/zaheer037/job-search --skill add-portal-zaheer037
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: add-portal
Source: https://github.com/zaheer037/job-search/tree/main/.agents/skills/add-portal
Command: npx skills add https://github.com/zaheer037/job-search --skill add-portal-zaheer037

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Adding support for a new job board to an agentic job-search workspace normally means hand-writing a scraper, guessing at URL patterns, and hoping the parsing works. This Skill turns that into a guided workflow: it investigates the portal's real endpoints, scaffolds a standardized search CLI, and verifies it against live data before registering anything. ## Core Features & Use Cases - Portal reconnaissance: Fetches the target job board to identify search endpoints, query parameters, result fields, robots.txt rules, and authentication requirements before writing any code. - Standardized scaffolding: Generates a complete skill directory (SKILL.md, url-reference.md, TypeScript CLI with search/detail commands, tests) following a strict portal-skill contract with zero runtime dependencies. - Mandatory live verification: Runs a real search query and detail fetch, iterating on parsers until results are clean, so no broken skill ever gets registered. - Use Case: A user in Australia wants to search Seek from their job-search workspace. They invoke /add-portal with the Seek URL, answer a few questions, and receive a tested seek-search skill that plugs into their existing /scrape workflow. ## Quick Start Ask the agent to run /add-portal with the URL of the job board you want to add, then answer the interview questions about market, language, and a test query.

Frequently Asked Questions about add-portal

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add a new job board to my job search workspace?▼

Invoke /add-portal with the portal's URL. The workflow interviews you for the market and a test query, investigates the site's search endpoints, scaffolds a search CLI, and verifies it with a live query before registering the skill.

How to build a job board scraper CLI with Bun and TypeScript?▼

The generated CLI uses plain Bun with fetch and regex parsing, zero runtime dependencies. It exposes search and detail commands with flags like --query, --jobage, --page, and --format, outputting a standard JSON shape with meta and results.

Does the portal generator work for job sites in any country?▼

Yes, the generator is country-agnostic and works for any portal in any market or language. The skills it produces are market-specific and stay in your fork, since upstream policy does not merge country-specific portal skills.

What happens if a job board requires login or blocks scraping?▼

The workflow stops for login-walled portals and suggests checking for an official API. If robots.txt or terms restrict access, it tells you plainly; if you proceed for personal use, the generated skill carries a prominent personal-use-only warning.

Why does the generated portal skill need a live test before registration?▼

Markup assumptions from reconnaissance routinely miss quirks that only appear in live responses. The mandatory test run verifies search returns real results and detail returns readable text, iterating on parsers until output is clean.

Can a generated portal skill use a paid fetching API for bot-protected sites?▼

Yes, when robots.txt permits access but bot protection blocks ordinary fetches. The token is read only from a SERVICE_API_TOKEN environment variable, never hardcoded, and per-call billing is disclosed before scaffolding.