company-context

分类: 工具与效率 | 上传者: katilevina | 下载: 0 | 版本: v1.0（最新）
Quick setup workflow for collecting company context upfront, including web research to trigger memories and enable smarter achievement gathering. Use when starting fresh, adding new companies, or updating existing company profiles before adding achievements.
更新日志: Source: GitHub https://github.com/katilevina/starry
目录结构

当前层级: .claude/skills/company-context/
📁 references/
- 📄 quick_setup_reference.md 7.4 KB
📄 SKILL.md 8.6 KB
SKILL.md

---
name: company-context
description: Quick setup workflow for collecting company context upfront, including web research to trigger memories and enable smarter achievement gathering. Use when starting fresh, adding new companies, or updating existing company profiles before adding achievements.
---

# Company Context

## Why Gather Context First?

### Benefits

**Enables smarter questions:**
- Instead of: "Who did you work with?"
- I ask: "Who from [Product, Engineering, Customer Success] did you collaborate with?"

**Faster achievement gathering:**
- Less clarification needed
- Context guides the conversation

**Better quality achievements:**
- Specific details from the start
- More accurate terminology

**Web research triggers memories:**
- "Oh right, I worked on that launch!"
- Helps date achievements more accurately

---

## Quick Setup Process

### Step 1: Check for Existing Data

**Search for:**
- Company profiles in `companies_i_worked/my_data/`
- Achievements in `achievements/my_data/`

**If data already exists:**

**Scenario A: Company profiles exist**
- Show list: "I found [X] company profiles already:"
- Ask: "Do you want to:"
  - Skip existing companies and add new ones?
  - Update existing profiles with new info?
  - Start fresh (overwrite all)?

**Scenario B: Only achievements exist (no company profiles)**
- Say: "I found [X] achievements but no company profiles yet."
- Ask: "Should I extract company context from your achievements automatically?"
  - If yes: Extract from achievements → show → confirm
  - Then fill gaps with targeted questions
  - If no: Proceed to manual collection

**Scenario C: Both exist**
- Say: "I found [X] company profiles and [Y] achievements."
- Explain: "Quick setup will update company profiles with new info, extract from achievements if gaps exist, add any new companies you mention"
- Confirm: "Proceed?"

### Step 2: Collect Company List

Ask:
- "How many companies have you worked for?"
- "List them briefly (most recent first)"
- If updating existing: focus on NEW companies or CHANGES

### Step 3: For Each Company, Ask

**Basic info:**
- Company name (for folder naming)
- **Employment dates** (YYYY-YYYY format) — REQUIRED for web research
- What do they do? (product/service, 1 sentence)
- Industry? (SaaS, E-commerce, Fintech, Healthcare, Consulting, Other)
- Company size? (Startup <50, Small 50-200, Mid 200-1000, Large 1000+)
- Business model? (B2B/B2C, subscription, marketplace, ads, etc)

**If profile exists:**
- Show current info → ask what to update

**If profile exists & achievements exist:**
- Show what I extracted from achievements → ask to confirm/add

### Step 4: Offer Web Research (PER COMPANY)

Ask: "Want me to search online for information about [Company Name] during [YYYY-YYYY]?"

**If yes:**
- Use web search (Perplexity or standard search)
- Look for company milestones during that period
- Extract relevant information that would help user remember achievements
- Store findings with `source: web_research` tag in company profile

**If no:**
- Proceed to next company or create profile with basic info

### Step 5: Create or Update Company Profiles

**If new:**
- Set `status: draft`
- Set `achievements_based_on: 0`
- Use template from company-profiles skill

**If updating:**
- Keep `achievements_based_on` count
- Update status
- Merge: existing + achievements + new input + web research

### Step 6: Update Index

- Add to or update `companies_i_worked/my_data/companies_index.md`
- Mark status appropriately (draft or complete)

---

## Web Research Feature

### Why Web Research Helps

**Triggers memories:**
- "Oh right, I worked on that launch!"
- Provides specific company events during user's tenure
- Adds rich context without manual effort

**Helps date achievements:**
- "When did you launch that feature?"
- "In 2020, they launched Product X — did you work on this?"

### What to Search For

When user provides employment dates (e.g., "2020-2023" for "Google"), search for:

**Company milestones:**
- Funding rounds
- Acquisitions
- IPO
- Major leadership changes

**Product launches:**
- New products during that period
- Feature releases
- Product updates

**Company growth:**
- Employee count changes
- Expansion to new markets
- Office openings

**News and events:**
- Awards
- Partnerships
- Major press coverage

**Industry trends:**
- Market shifts during that period
- Competitor moves
- Regulatory changes

### Example Search Queries

```
"[Company Name] [YYYY-YYYY] product launches"
"[Company Name] funding rounds 2020 2021 2022"
"[Company Name] during [year range] news"
"[Company Name] [year] milestones"
"[Company Name] employee growth [years]"
```

### How to Use Findings

After collecting web research:

**1. Store findings:**
- Add to company profile with `source: web_research` tag
- Include date: `*(source: web_research, date: YYYY)*`
- Create section: "Web Research Findings (Optional)"

**2. Present to user:**
```
I found some interesting information about Acme Corp online during your employment (2020-2023):

- In 2021, they raised Series B funding and grew from 50 to 150 employees
- Launched their flagship product "AcmePro" in 2022
- Expanded to European market in early 2023

This will help me ask better questions! For example: "Did you work on the AcmePro launch in 2022?"
```

**3. Use for targeted questions:**
- "In 2020, Google launched Google Workspace — did you work on this?"
- "During your tenure, Acme Corp grew from 50 to 200 employees. How did this affect your role?"
- "I see they launched [Product] in [Year]. Was your achievement related to this?"

### Sourcing: Always Mark Web Research

**DO:**
- Always mark with `*(source: web_research, date: YYYY)*`
- Distinguish between user-provided info and web findings
- Allow user to correct or reject findings

**DON'T:**
- Don't present web research as fact without source tag
- Don't assume user worked on everything mentioned
- Don't overwhelm with too much detail

### When to Skip Web Research

**Skip if:**
- Small companies with no online presence
- Very old employment dates (pre-internet or limited archives)
- User declines the offer
- Company operates in stealth mode
- Employment dates are missing (can't search effectively)

---

See [Quick Setup Reference](references/quick_setup_reference.md) for detailed workflow examples.

## How Context Improves Achievement Gathering

### Before Context (Generic Questions)

- "What was the situation?"
- "Who did you work with?"
- "What tools did you use?"
- "What were the constraints?"

### After Context (Smart Questions)

**Using industry knowledge:**
- "What was the situation at [SaaS company]? Were you working on [product launches, customer retention, feature development]?"

**Using company info:**
- "Who from [Product, Engineering, Customer Success] did you collaborate with?"

**Using typical tools:**
- "Did you use [Jira, Figma, Amplitude] for this?"

**Using web research:**
- "In 2020, [Company] launched [Product X]. Did your achievement relate to this?"
- "During [Year], the company grew from X to Y employees. How did this affect your work?"
- "Given that [market condition] during your tenure, what challenges did you face?"

---

## Data Structure

### What Gets Collected

**For each company:**

**Basic info (always):**
- Company name
- Employment dates (YYYY-YYYY)
- Industry
- Size
- Business model

**Optional (web research):**
- Company milestones during tenure
- Product launches
- Growth metrics
- Major events

### Profile Status

**Draft:**
- `status: draft`
- `achievements_based_on: 0`
- Basic info only
- May have web research findings

**Complete:**
- `status: complete`
- `achievements_based_on: 3+` (or however many)
- Built from achievements
- Has web research (optional)
- Culture/context sections filled

---

## Quality Checklist

Before finishing quick setup:

* [ ] Checked for existing data first (companies + achievements)
* [ ] If achievements exist, showed what I extracted before asking questions
* [ ] Collected employment dates (REQUIRED for web research)
* [ ] Offered web research for each company
* [ ] Marked web-researched info with `source: web_research` tag
* [ ] Created or updated company profiles
* [ ] Updated companies_index.md
* [ ] Confirmed next steps with user

---

## Integration with Other Skills

### With STARR Achievements

**When user runs `/add-achievement`:**
- Check if company profile exists
- If yes: Load context (industry, size, business model, typical stakeholders, tools)
- Check for web research findings
- Use context + web research to ask smart questions

### With Company Profiles

**When user runs `/add-company`:**
- Check for achievements first
- If achievements exist, extract from them
- Fill gaps with targeted questions
- Web research findings already in profile from quick setup
登录后下载/点赞/收藏 ❤ 4 | ★ 0
company-context

目录结构

SKILL.md

举报内容

提示
