Perform multi-engine web searches and structured OSINT investigations via HTTP endpoints.
npx skills add https://github.com/erfianugrah/dotfiles --skill research-erfianugrah
Please help me install this Agent Skill. Skill: research Source: https://github.com/erfianugrah/dotfiles/tree/main/.pi/agent/skills/research Command: npx skills add https://github.com/erfianugrah/dotfiles --skill research-erfianugrah
It eliminates slow, manual OSINT workflows by providing a single set of HTTP endpoints for multi-engine search, clean content extraction, and structured investigation across domains, IPs, emails, usernames, URLs, phones, and reputations.
Use the OSINT stack to investigate a domain by sending an authenticated request to the local OSINT service: run curl with a POST to /investigate/domain using JSON with the target domain and desired mode.
High-intent search queries and answers about installing and using this skill.
Domain investigation and OSINT lookups are automated by sending authenticated HTTP POST requests to an OSINT service, which coordinates DNS discovery, IP correlation, and username scans across many platforms. You use JSON payloads to specify the target and desired mode.
Yes, clean content extraction from JavaScript-heavy single page applications is handled by a crawler extractor service that strips boilerplate into Markdown. It uses a fast extraction path first and falls back to Playwright to render dynamic content.
You need SearXNG running on port 8888 for multi-engine search, a crawler extractor on port 8889 for webpage fetching, and an OSINT service on port 8890 for structured investigations. All endpoints require bearer authentication.
Threat intelligence gathering supports long-running asynchronous jobs by submitting targets to endpoints like urlscan.io lookups and polling the /jobs/{id} path. This allows you to retrieve structured results once the background investigation completes.
CVE lookups and IP reputation checks are performed by querying the OSINT service's /investigate/* endpoints, which correlate IP ports with NVD CVE data and fetch VirusTotal threat reputation scores. You submit the target indicator via JSON in an authenticated request.
The best way to aggregate web search results is to query a local SearXNG instance on port 8888, which merges results across 7+ search engines. This provides a single endpoint to discover relevant sources and leads for research.
mrudExpose local services over secure self-hosted tunnels with public and private sharing.
Devsoul2026Extract structured data from LLM responses with Pydantic validation and automatic retries.
workromancerFetch web URLs and extract all text content.
LeoWSY-hashblueFetch CVE risk scores from the National Vulnerability Database when CVE IDs are detected.
