AI agent skill · by @AlphaGBM

Alphagbm Company Profile

Build and maintain company research profiles on AlphaGBM — auto-generated from fundamentals, PE/PB Band history, financial red flags, and event radar. Each profile is one user+ticker record that the system refreshes on schedule. Use when: creating a watchlist of companies to track, pulling up a saved research file, refreshing a profile's market data, or checking PE/PB bands. Triggers on: "add AAPL to my knowledge base", "show my profile for NVDA", "refresh my TSLA profile", "list my tracked companies", "PE band for META", "what's in my research brain", "创建公司档案", "我的投研档案".

Archived 2026-10-23 · Source updated 2026-09-14 · License: MIT. Check the original repository for the latest version.

SKILL.md instructions

# AlphaGBM Company Profile

Build and manage company research profiles in a user's private knowledge base. Each profile captures fundamentals (PE/PB), 8-year valuation bands, financial red flags, and recent events — auto-refreshed on a schedule.

## When to use

- User wants to track a company in their personal research workspace
- User asks to list / view / delete saved companies
- User asks for PE or PB historical band of a ticker
- User asks to refresh a stale profile
- User mentions "知识库" / "投研档案" / "research brain" / "knowledge base"

## Prerequisites

- **API Key**: stored in env `ALPHAGBM_API_KEY` (format `agbm_xxxx…`).
- **Base URL**: default `https://alphagbm.zeabur.app`. Override with env `ALPHAGBM_BASE_URL`.
- If the user has no key, direct them to register at <https://alphagbm.com> and create one at `/api-keys`.
- **Tier requirement**: Free tier = 1 profile, Plus = 10, Pro = 50. Create endpoint returns 403 with `upgrade_required: true` when the cap is hit.

## API Endpoints

All endpoints require `Authorization: Bearer $ALPHAGBM_API_KEY`.

### 1. List profiles

```
GET /api/research/profiles?page=1&per_page=20
```

**Response:**
```json
{
  "success": true,
  "profiles": [{ "ticker": "AAPL", "company_name": "...", "current_price": 261.0, ... }],
  "total": 3,
  "page": 1,
  "per_page": 20
}
```

### 2. Get profile detail (includes thesis if one exists)

```
GET /api/research/profiles/<TICKER>
```

Returns the full profile + an embedded `thesis` field (null if no thesis yet). See **Response schema** below for field list. Returns 404 if the ticker isn't in the user's knowledge base.

### 3. Create profile

```
POST /api/research/profiles
Content-Type: application/json

{"ticker": "AAPL"}
```

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `ticker` | string | yes | Stock ticker (US / HK / A-share), case-insensitive |

**Behavior:** Pulls fundamentals via the data provider, computes red flags and event radar, persists the profile. If a profile already exists for this user+ticker, it's updated in place (idempotent).

**Tier limit response (403):**
```json
{
  "success": false,
  "error": "Profile limit reached. Upgrade to Plus for 10 profiles.",
  "current": 1,
  "max": 1,
  "upgrade_required": true
}
```

### 4. Delete (archive) profile

```
DELETE /api/research/profiles/<TICKER>
```

Soft-deletes by flipping `status` to `archived`. Returns 404 if not found.

### 5. Refresh profile data

```
POST /api/research/profiles/<TICKER>/refresh
```

Pulls fresh market data, recomputes red flags and events. Use when the user says "refresh my profile" or the `last_updated_at` is stale (> 7d old).

### 6. PE/PB Band data (cached 24h)

```
GET /api/research/profiles/<TICKER>/band
```

Returns 8-year PE/PB history for building the band chart. Can be called **without** the ticker being in the user's knowledge base — it's a read-only market data endpoint.

**Response:**
```json
{
  "success": true,
  "ticker": "AAPL",
  "pe_history": [{"date": "2017-04", "pe": 16.2}, ...],
  "pb_history": [...],
  "current_pe_percentile": 0.82,
  "current_pb_percentile": 0.75
}
```

## Response schema — full profile

```
{
  id, ticker, company_name, market,          // market = US | HK | CN
  current_price, pe_ratio, pb_ratio,
  pe_band_data,                              // 8yr history, same shape as /band endpoint
  financial_red_flags,                       // [{rule_id, severity: "high|med|low", message}]
  event_radar,                               // [{event_type, timestamp, headline}]
  ai_profile_summary,                        // markdown, ~500 chars
  status,                                    // "active" | "archived"
  last_viewed_at, last_updated_at, created_at
}
```

## Typical Workflow

```
1. User: "Add NVDA to my research brain"
   → POST /api/research/profiles {"ticker": "NVDA"}
   → Present: "Added NVDA. Current PE 45, 2 red flags, PE at 85th percentile of 8yr range."

2. User: "What's in my knowledge base?"
   → GET /api/research/profiles
   → Present table: ticker · company · PE · last_updated · red flag count

3. User: "Show me my AAPL profile"
   → GET /api/research/profiles/AAPL
   → Present: summary, PE/PB bands, red flags list, event radar, linked thesis (if any)

4. User: "Refresh my TSLA profile"
   → POST /api/research/profiles/TSLA/refresh
```

## Tier Limits

| Tier | Max profiles |
|------|-------------|
| Free | 1 |
| Plus | 10 |
| Pro  | 50 |

When a create hits the limit, the API returns `upgrade_required: true`. Surface this to the user with a prompt to upgrade at `/pricing`.

## Output Formatting Tips

When presenting a profile to the user, highlight:
1. **Ticker + company name** and market flag (US / HK / CN)
2. **Current price + PE / PB** with percentile context ("PE 32, 85th percentile of 8yr range → rich")
3. **Red flags** — group by severity, show top 3
4. **Event radar** — most recent 3-5 events with dates
5. **Linked thesis** — if present, one-line buy reason + exit trigger summary
6. **Staleness** — if `last_updated_at` > 7d old, suggest a refresh

## Related Skills

- **alphagbm-investment-thesis** — Attach buy thesis + exit triggers to a profile
- **alphagbm-health-check** — Detect stale / drifted profiles across the user's workspace
- **alphagbm-stock-analysis** — One-off deep analysis (not persisted to the knowledge base)
- **alphagbm-theme-research** — Group profiles into themes (AI infra, HK dividend, etc.)

---

*Powered by [AlphaGBM](https://alphagbm.com) — Real-data options & research intelligence for traders and AI agents. 10K+ users.*