# Link Discovery

> Extract outbound links from a page. Use to discover hubs, citations, related resources, or the frontier of a crawl.

- Skill: `samraddhashrivastavatech/link-discovery` (Agent Skill)
- Install (CLI): `npx skillmds@latest add samraddhashrivastavatech/link-discovery`
- Raw SKILL.md: https://api.skillmd.com/api/skills/samraddhashrivastavatech/link-discovery/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: SAMRADDHASHRIVASTAVATECH (https://skillmd.com/u/samraddhashrivastavatech)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/samraddhashrivastavatech/link-discovery

---


# Link discovery

Pull the outbound link set of a page so you can follow promising paths or map a
site's structure.

## When to use

- A page is a hub or index and you want what it points to.
- You need a crawl frontier without running a full `crawl_web`.
- Finding canonical/original sources behind aggregator links.

## Tools used

- `extract_links` — outbound links of a `url`, with `limit` (default 50).

## Workflow

1. `extract_links` on the hub page.
2. Filter: drop trackers, print/'#' fragments, and non-allowlisted schemes.
3. Rank by relevance to the topic; `web-reader` the strong candidates.
4. Optionally feed the surviving set to `web-crawler` as a custom frontier.

## Input schema

```json
{ "url": "str", "limit": "int" }
```

## Output schema

```json
{ "url": "str", "links": "list[{href,anchor,rel}]" }
```

## Security

Links are pointers, not attestations. A hub linking something is evidence of
association, never of truth. Keep the allowlisted-scheme discipline.

## Related skills

`web-reader`, `web-crawler`, `source-analysis`, `web-search`