competitor-analysis
/competitor-analysis
Deep extract top-5 SERP competitors via Tavily 5-tier waterfall. For each: title/meta/H2 structure/word count/schema types/content gap. Triggered by phase-research after serp-analysis.
- Author
- xuanranl ↗
- Repository
- loamwright-seo-skill ↗
- Licence
- Apache-2.0
- Last repository change
- 17.08.2026
Assessment and security
- Files in this package
- 1
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 1
No signals detected. Repository changed on 17.08.2026, without references to retired models or APIs.
Apache-2.0, otwarta: można używać i zmieniać, z zachowaniem licencji i informacji o zmianach.
What it does and when to use it
Deep extract top-5 SERP competitors via Tavily 5-tier waterfall. For each: title/meta/H2 structure/word count/schema types/content gap. Triggered by phase-research after serp-analysis.
Original author description
Deep extract top-5 SERP competitors via Tavily 5-tier waterfall. For each: title/meta/H2 structure/word count/schema types/content gap. Triggered by phase-research after serp-analysis.
How it works
- Inputs
- Workflow
- Output schema fragment (added to research.json)
- Cost
SKILL.md preview
---
name: competitor-analysis
description: Deep extract top-5 SERP competitors via Tavily 5-tier waterfall. For each: title/meta/H2 structure/word count/schema types/content gap. Triggered by phase-research after serp-analysis.
allowed-tools: [Read, Write, Bash, Task]
---
# Competitor Analysis
## Inputs
- `research.json` (top-10 SERP URLs from keyword-research)
- Own domain (excluded from competitor list)
## Workflow
1. Filter top-10 to top-5 (exclude own domain + aggregators like Amazon if not aggregator-focused)
2. For each: `python -m scripts.fetch.multi_tier_fetch {url}` (use 5-tier waterfall)
3. `python -m scripts.fetch.parse_html {html_file} --json` to extract title/meta/H2/schema
4. LLM analyzes content_gap: what did THIS competitor cover shallowly vs deeply?
5. Save to `workspace/{task}/research/competitors/{i}.json`
6. **Competitor-citation candidate review (Rule 8).** The SERP competitor domains
you just identified are, by definition, peers ("同行"). Diff them against the
project's enforced blocklist and surface any NEW ones for the operator to
approve — never auto-block (avoids false-positives on neutral sites that merely
rank, e.g. Wikipedia/.gov):
```bash
python -m scripts._core.competitor_domains --task {task_id} --json # current blocklist + enabled?
```
For each competitor domain NOT already in `do_not_cite_domains`, append it to
`workspace/{task}/competitor-candidates.json` as `{"domain": ..., "source": "serp",
"keyword": ...}` and note it in the research handoff so the operator can promote
genuine competitors into `business-context.json :: citation_source_policy.do_not_cite_domains`.
These candidates are NOT enforced until promoted.
## Output schema fragment (added to research.json)
```json
"competitor_titles": [
{
"title": "Best Fishing Rods 2026 — Saltwater Guide",
"url": "https://competitor.com/...",
"domain": "competitor.com",
"word_count_estimate": 3200,
"content_gap": "Missing pricing comparison; weak on saltwater-specific testing",
"schema_types": ["Article", "BlogPosting"],
"freshness_days": 42
}
]
```
## Cost
- 5 × Tavily advanced extract = 5 credits ($0.04)
- 1 × Claude Opus synthesis = $0.05
- Total: ~$0.10
Versions
| Repository | Score | Licence |
|---|---|---|
| This page · xuanranl/loamwright-seo-skill | 100 | Apache-2.0 |
| aaron-he-zhu/aaron-marketing-skills | 100 | Apache-2.0 |
| opencue/cuecards | 100 | MIT |
| opencue/cuecards | 100 | MIT |
| AmmarBabbar/seo-skills | 100 | MIT |
| Kuan-AI-com/agent-skills | 100 | MIT |
| Zoominfo/zoominfo-mcp-plugin | 100 | MIT |
| lingegrace/claude-research-skills | 100 | MIT |
| OpenClaudia/openclaudia-skills | 100 | MIT |
Database snapshot: 13.11.2026 · package version 0e85405ff6cc (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 100phase-research
Run Phase 1 Research only — keyword research, SERP analysis, competitor analysis, content gap analysis, surface targeting. Use when user wants research-only output without writing an article. Triggered by /seo-blog research, "research this keyword", "what's the SERP for X",… - 100seo-competitor-analysis
Deep SEO competitor analysis — keyword mapping, backlink profiling, content strategy audit, SERP share analysis, and technical SEO comparison. Use when user asks to 'analyze competitor SEO', 'competitor keyword analysis', 'backlink comparison', 'SEO competitive intelligence',… - 90competitor-research
When the user wants to analyze competitors for SEO, content, backlinks, or positioning. Also use when the user mentions "competitor analysis," "competitor research," "competitor keywords," "competitor backlinks," "link gap," "content gap," "competitor content," "competitive… - 90competitor-research
When the user wants to analyze competitors for SEO, content, backlinks, or positioning. Also use when the user mentions "competitor analysis," "competitor research," "competitor keywords," "competitor backlinks," "link gap," "content gap," "competitor content," "competitive… - 100serp-extract
When the user wants to capture SERP evidence, search result snapshots, competitor URLs, or SERP features for specific keywords without doing SEO analysis. - 100keyword-research
Deep keyword research via Tavily SERP + Crossref + AI-search probe. Builds research.json with intent, semantic clusters, PAA, competitor titles, SERP features, AI engine findings, content gaps. Triggered by phase-research orchestrator or directly when user requests keyword…