Linkedin Reader

# LinkedIn Skill (Read-Only)

Reads LinkedIn for financial research using [opencli](https://github.com/jackwener/opencli), a universal CLI tool that bridges web services to the terminal via browser session reuse.

**This skill is read-only.** It is designed for financial research: reading professional commentary on markets, monitoring analyst posts, searching finance/trading jobs, and tracking professional sentiment. It does NOT support posting, liking, commenting, connecting, messaging, or any write operations.

**Important**: opencli reuses your existing Chrome login session — no API keys or cookie extraction needed. Just be logged into linkedin.com in Chrome and have the Browser Bridge extension installed.

---

## Step 1: Ensure opencli Is Installed and Ready

**Current environment status:**

```
!`(command -v opencli && opencli doctor 2>&1 | head -5 && echo "READY" || echo "SETUP_NEEDED") 2>/dev/null || echo "NOT_INSTALLED"`
```

If the status above shows `READY`, skip to Step 2. If `NOT_INSTALLED`, install first:

```bash
# Install opencli globally
npm install -g @jackwener/opencli
```

If `SETUP_NEEDED`, guide the user through setup:

### Setup

opencli requires Node.js >= 21 and a Chrome browser with the Browser Bridge extension:

1. **Install the Browser Bridge extension:**
   - Download the latest `opencli-extension-v{version}.zip` from the [GitHub Releases page](https://github.com/jackwener/opencli/releases)
   - Unzip it, open `chrome://extensions` in Chrome, and enable **Developer mode**
   - Click **Load unpacked** and select the unzipped folder
2. **Login to linkedin.com** in Chrome — opencli reuses your existing browser session
3. **Verify connectivity:**

```bash
opencli doctor
```

This auto-starts the daemon, verifies the extension is connected, and checks session health.

### Common setup issues

| Symptom | Fix |
|---------|-----|
| `Extension not connected` | Install Browser Bridge extension in Chrome and ensure it's enabled |
| `Daemon not running` | Run `opencli doctor` — it auto-starts the daemon |
| `No session for linkedin.com` | Login to linkedin.com in Chrome, then retry |
| `AuthRequiredError` | LinkedIn session expired — refresh linkedin.com in Chrome and log in again |

---

## Step 2: Identify What the User Needs

Match the user's request to one of the read commands below, then use the corresponding command from `references/commands.md`.

| User Request | Command | Key Flags |
|---|---|---|
| Setup check | `opencli doctor` | — |
| Home feed / posts | `opencli linkedin timeline` | `--limit N` (default 20, max 100) |
| Search for jobs | `opencli linkedin search "QUERY"` | `--location`, `--limit N` (default 10, max 100), `--details` |
| Finance job search | `opencli linkedin search "QUERY"` | `--experience-level`, `--job-type`, `--remote`, `--company`, `--date-posted`, `--start` |

---

## Step 3: Execute the Command

### General pattern

```bash
# Read LinkedIn feed posts
opencli linkedin timeline --limit 20 -f json

# Search for finance/trading jobs
opencli linkedin search "quantitative analyst" --limit 10 -f json
opencli linkedin search "portfolio manager" --location "New York" --limit 15 -f json

# Detailed job listings with descriptions
opencli linkedin search "financial analyst" --details --limit 10 -f json
```

### Key rules

1. **Check setup first** — run `opencli doctor` before any other command if unsure about connectivity
2. **Use `-f json` or `-f yaml`** for structured output when processing data programmatically
3. **Use `-f csv`** when the user wants spreadsheet-compatible output
4. **Use `--limit N`** to control result count — start with 10-20 unless the user asks for more
5. **For job search, use filters** — `--location`, `--experience-level`, `--job-type`, `--remote`, `--date-posted` to narrow results
6. **NEVER execute write operations** — this skill is read-only; do not post, like, comment, connect, message, or apply to jobs

### Output format flag (`-f`)

| Format | Flag | Best for |
|---|---|---|
| Table | `-f table` (default) | Human-readable terminal output |
| JSON | `-f json` | Programmatic processing, LLM context |
| YAML | `-f yaml` | Structured output, readable |
| Markdown | `-f md` | Documentation, reports |
| CSV | `-f csv` | Spreadsheet export |

### Output columns

**Timeline** posts include: `rank`, `author`, `author_url`, `headline`, `text`, `posted_at`, `reactions`, `comments`, `url`.

**Job search** results include: `rank`, `title`, `company`, `location`, `listed`, `salary`, `url`. With `--details`: also `description`, `apply_url`.

---

## Step 4: Present the Results

After fetching data, present it clearly for financial research:

1. **Summarize key content** — highlight the most relevant posts or jobs for the user's research
2. **Include attribution** — show author name, headline, post text, and engagement (reactions, comments)
3. **Provide URLs** when the user might want to read the full post or job listing
4. **For feed posts**, highlight market commentary, analyst takes, earnings reactions, and professional sentiment
5. **For job search results**, present title, company, location, salary (when available), and posting date
6. **Flag sentiment** — note bullish/bearish professional sentiment, consensus vs contrarian views
7. **Treat sessions as private** — never expose browser session details

---

## Step 5: Diagnostics

If something isn't working, run:

```bash
opencli doctor
```

This checks daemon status, extension connectivity, and browser session health.

---

## Error Reference

| Error | Cause | Fix |
|-------|-------|-----|
| `Extension not connected` | Browser Bridge not installed/enabled | Install extension and enable it in Chrome |
| `No session` | Not logged into linkedin.com | Login to linkedin.com in Chrome |
| `AuthRequiredError` | LinkedIn login wall detected | Refresh linkedin.com and log in again |
| `EmptyResultError` | No results found for query | Broaden search terms or check feed has content |
| Rate limited | Too many requests | Wait a few minutes, then retry |

---

## Reference Files

- `references/commands.md` — Complete read command reference with all flags, research workflows, and usage examples

Read the reference file when you need exact command syntax, research workflow patterns, or output details.