Competitve-Intelligence-CLI

分类: 工具与效率 | 上传者: qb-harshit | 下载: 0 | 版本: v1.0（最新）
您可以使用“cintel”——一个竞争情报工具包。当用户询问竞争对手、竞争分析、战斗卡、定位、定价情报、功能比较或任何与了解另一家公司正在做什么相关的信息时，请使用此工具。
更新日志: Source: GitHub https://github.com/qb-harshit/Competitve-Intelligence-CLI
目录结构

当前层级: tree/main/competitive_intel/data/
📄 prompt_library.md 7.3 KB
📄 skill.md 7.5 KB
SKILL.md

# Competitive Intelligence Harness

You have access to `cintel` — a competitive intelligence toolkit. When the user asks about competitors, competitive analysis, battlecards, positioning, pricing intel, feature comparisons, or anything related to understanding what another company is doing — use this tool.

## Trigger Phrases

Activate when the user says anything like:
- "do competitive intelligence on [company/url]"
- "analyze [company] as a competitor"
- "scrape [company]'s website"
- "what does [company] offer / what are they doing"
- "compare [company] vs [company]"
- "check [company]'s pricing"
- "build a battlecard for [company]"
- "find gaps in [company]'s product"
- "what's [company] weak at"
- "SEO analysis for [url]"
- "competitive analysis", "competitor research", "market intelligence"
- "what changed on [company]'s site"
- "analyze [company]'s docs / help center / documentation"
- "what is [company] hiring for"

## Available Commands

```bash
cintel scrape homepage "<company>" "<url>"
cintel scrape sitemap "<company>" "<sitemap_url>" --keywords "features,pricing,docs,customers,api,faq,help" --category all
cintel scrape pricing "<company>" "<url>"
cintel scrape seo "<company>" "<url>"
cintel list
cintel view "<company>"
cintel view "<company>" --section homepage
cintel view "<company>" --section features
cintel view "<company>" --section pricing_data
cintel view "<company>" --section seo_data
```

Scraping takes 30-120 seconds per page. Use timeout 300000ms for all scrape commands.

## Full Competitive Intelligence Workflow

When the user asks you to do competitive intelligence on a company, run this entire workflow automatically. Don't stop after the homepage — go deep.

### Phase 1: Scrape everything

**Step 1 — Homepage**
```bash
cintel scrape homepage "<company>" "<url>"
```

**Step 2 — Sitemap discovery and bulk scraping**
Try `<domain>/sitemap.xml`. If it exists, scrape broadly:
```bash
cintel scrape sitemap "<company>" "<domain>/sitemap.xml" --keywords "features,pricing,docs,documentation,api,customers,case-study,help,faq,integrations,security,changelog,release-notes" --category all
```
This auto-categorizes URLs into: features, products, pricing, customers, faq, api, documentation, other.

**Step 3 — Pricing page**
```bash
cintel scrape pricing "<company>" "<domain>/pricing"
```

**Step 4 — SEO analysis**
```bash
cintel scrape seo "<company>" "<url>"
```

### Phase 2: Read and cross-reference ALL data

```bash
cintel view "<company>" --section homepage
cintel view "<company>" --section features
cintel view "<company>" --section pricing_data
cintel view "<company>" --section seo_data
```

Read the full JSON at `data/companies/<company>_data.json` for complete content.

### Phase 3: Deep analysis (this is where the real intel comes from)

Don't just summarize what you scraped. Analyze it through these lenses:

#### 1. Marketing vs Documentation Gap
This is the highest-value analysis. Compare what the homepage/marketing pages claim vs what the docs/help center pages reveal:
- Features marketed as available that docs show are "beta" or "preview"
- "Easy to use" claims contradicted by complex implementation docs
- Rate limits, quotas, or restrictions buried in technical docs but hidden from marketing
- Deprecated features still promoted on the marketing site
- Known issues or workarounds documented in help center

#### 2. Feature Status Mapping
Map every feature you found to a status:
- **GA** — generally available, documented, marketed
- **Beta/Preview** — mentioned in docs but flagged as not production-ready
- **Coming Soon** — on roadmap pages or mentioned as future
- **Deprecated** — still exists but being sunset
- **Hidden** — exists in docs but not on marketing site (could be powerful or could be abandoned)

#### 3. Positioning Deconstruction
- What exact words do they use to describe themselves? (the H1, the meta description, the first paragraph)
- What "role" do they claim to replace? ("We're your AI SDR" vs "sales automation platform")
- What's their primary CTA? ("Book a Demo" = enterprise sales motion, "Try Free" = PLG motion, "Contact Sales" = high ACV)
- Who are their listed customers? What verticals/segments do they target?
- What trust signals do they lean on? (SOC2, ISO, GDPR, customer logos, funding)

#### 4. Pricing Model Deconstruction
- Per user? Per seat? Usage-based? Flat rate? Contact sales?
- How many tiers? What's gated behind enterprise?
- Is there a free tier or free trial? How long?
- What's the cheapest entry point vs the enterprise ceiling?
- Any "starting at" language that hides the real price?

#### 5. Anomaly Detection
Look for things that don't fit or signal something interesting:
- Job postings (if mentioned on the site) — what roles are they hiring? (10 ML engineers = AI pivot, enterprise sales reps = upmarket move, developer advocates = PLG push)
- Changelog or release notes — what shipped recently? What's the velocity?
- Help center volume — lots of troubleshooting articles for a feature = it's buggy
- Missing pages — no security page? No compliance certs? No customer stories? That's a gap.
- Language shifts — "startup" language vs "enterprise" language reveals who they're selling to

#### 6. Ecosystem and Integration Map
- What integrations do they list?
- What's native vs partner-built?
- What APIs do they expose? REST? GraphQL? Webhooks?
- What developer resources exist? SDKs, docs quality, community

### Phase 4: Produce the report

Structure the output as:

#### Company Overview
What they do, who they serve, how they position themselves. Use their exact words.

#### Product & Feature Map
All features found, organized by category. Include status (GA/beta/deprecated). Note which features appear in docs but not on marketing pages.

#### Pricing Intelligence
Full pricing breakdown: model, tiers, amounts, currencies, billing, trials, enterprise signals.

#### Documentation & Help Center Findings
What the docs reveal that marketing doesn't: limitations, restrictions, complexity, known issues.

#### SEO & Digital Presence
Score, trust signals, content strategy, technical SEO health.

#### Anomalies & Signals
Anything that stands out: gaps, inconsistencies, strategic signals from hiring/changelog/missing content.

#### Strengths (be specific, cite evidence from scraped content)

#### Weaknesses & Gaps (be specific, cite evidence)

#### Actionable Output for Sales
- 3-5 key talking points when competing against this company
- Top objections a buyer might raise and how to handle them
- "Trap questions" that position your product favorably
- Feature gaps to exploit in a competitive deal

Save the report to `data/companies/<company>/full_report.md`.

## Multi-Company Comparison

When comparing competitors:
1. Run the full workflow above for each company
2. Read all company data files
3. Produce:

**Feature Matrix** — Side-by-side table of capabilities with status (has it / doesn't / partial / beta)

**Pricing Comparison** — Tier-by-tier, model-by-model breakdown

**Positioning Map** — How each describes themselves, who they target, what role they claim to replace

**Strengths vs Weaknesses Grid** — Top 3 each per company with evidence

**Competitive Battlecard** — For each competitor pair:
- When we win (scenarios, segments, use cases)
- When they win (be honest)
- Key differentiators
- Objection handling
- Trap questions for discovery calls

Save to `data/companies/comparison_<company1>_vs_<company2>.md`.

## Data Location

All data: `data/companies/<company>_data.json`
Reports: `data/companies/<company>/<report>.md`
Prompt templates: `prompts/prompt_library.md`
登录后下载/点赞/收藏 ❤ 38 | ★ 0
Competitve-Intelligence-CLI

目录结构

SKILL.md

举报内容

提示
