# Source Analysis

> Classify a source's authority, type, and freshness. Use to decide how much weight a specific URL deserves before it enters a verdict.

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

---


# Source analysis

Profile where a source sits: how authoritative its domain/type is, what the
source is, and how fresh it is.

## When to use

- A URL is about to be cited or weigh on a verdict.
- Deciding between conflicting sources (reliability triage).
- Auditing a result set before evidence extraction.

## Tools used

- `source_analysis` — analyze a URL's authority/type/freshness.
- `get_skill` — pull this skill's own definition (discovery/self-doc).

## Authority/type/freshness axes

- **Authority** — primary/official vs. secondary vs. aggregator vs. self-published;
  domain reputation.
- **Type** — gov, academic, vendor, documentation, news, forum, social, blog.
- **Freshness** — dated content vs. evergreen; last-known update.

## Workflow

1. For each candidate URL, `source_analysis`.
2. Record the classification next to the URL in your evidence set.
3. Let type + freshness inform weight, but never let it fabricate evidence —
   a low-authority source can still be the only one, say so.
4. Re-classify if you later find the true author/publication center.

## Input schema

```json
{ "url": "str" }
```

## Output schema

```json
{ "url": "str", "authority": "str", "type": "str", "freshness": "str", "recommendation": "str" }
```

## Security

Classification is an aid to judgment, not proof of truth. Never cite or assert
a source's content as fact purely on its "authoritative" label.

## Related skills

`web-reader`, `web-search`, `claim-verification`, `comparison`,
`deep-research`