radar · git:20260731.24e4af6 · 2026-07-31 · sha256 cb46e3c5b9eb9dc2
radar git:20260731.24e4af6F
Immutable. This exact content never changes and is served at /api/v1/blob/cb46e3c5b9eb9dc2.
No tool to install first
claude plugin marketplace add https://markdownregistry.com/mp/a/art_rt2gorj3tucvqv7a/cb46e3c5b9eb9dc24e59703ea3e082d5a0a8b4459a4633fda3b0f96574efb811.json
Then /plugin install radar@markdownregistry-rt2gorj3tucvqv7a in Claude Code. It installs this exact version, the one graded F here, pinned to the commit it was graded at, and leaves it there.
---
name: radar
description: Registers the user on job platforms, configures alerts with profile keywords, creates a Gmail filter to route alerts to a Job Alerts folder.
trigger: radar
---
# Radar
## Trigger
**Keyword: `radar`**
The user says `radar` (or variants: "set up alerts", "register on platforms", "new platforms") and the registration + alert configuration + Gmail filter flow is triggered.
## Purpose
Passive sourcing: platforms bring opportunities to the user without having to search manually. Alerts arrive in a `Job Alerts` folder in Gmail, and the `news` skill consumes them when the user says `news`.
## Flow
### 0. Pre-flight
- [ ] Verify active Gmail session. If session closed → open browser with wrapper (see AGENTS.md "Browser session"): `node scripts/browser.js open <url> --headed` (Gold Rule 5)
- [ ] **Browser:** always use `node scripts/browser.js` for open/close/goto. See AGENTS.md "Browser session" for details. Never call `playwright-cli open` directly, never open Chrome directly
- [ ] Read `PLATFORMS.md` "Alert Tracking" section to see which platforms need configuration
- [ ] Read `PROFILE.md` to get keywords, seniority, location, preferences
- [ ] Load active preferences (see `memory` skill):
```bash
node scripts/db.js "SELECT category, key, value, confidence, source FROM preferences WHERE user_id = 1 AND status = 'active' ORDER BY category, key"
```
- [ ] Load target companies (40 companies: 19 LATAM + 21 Argentina, each `{name, url, sector}`). These are priority targets for alert configuration and platform prioritization:
```bash
node scripts/db.js "SELECT data->'target_companies' AS target_companies FROM users WHERE id = 1"
```
### 1. Register on platforms
For each unconfigured platform (column "Profile" = "—"):
- [ ] Navigate to the platform
- [ ] **Login with Google** when possible (preferred). If no Google login → open browser with wrapper (`node scripts/browser.js open <url> --headed`) and ask user to login manually (Gold Rule 5)
- [ ] Complete minimum profile:
- First name, last name
- Title / headline (from PROFILE.md)
- Location (from PROFILE.md)
- Seniority (from PROFILE.md)
- Upload CV (CV path in DB or `.env`)
- Preferences: remote, full-time, USD salary range
- [ ] Mark "Profile" = "✅" in PLATFORMS.md
### 2. Configure alerts
For each platform with complete profile but no alerts:
- [ ] Create alert with profile keywords:
- `AI Architect`, `Engineering Manager`, `AI Strategy`, `AI Implementation`
- `LLM`, `Agent-First`, `AI Workflow`, `SDLC AI`
- `Technical Lead`, `Staff Engineer`, `Platform Engineer`
- [ ] If the platform supports company-specific alerts or saved searches, configure alerts for target companies from the DB (40 companies: Nubank, dLocal, Mercado Libre, Globant, Ualá, BBVA, Santander Tecnología, Auth0, etc.). Group by sector if the platform allows it (Fintech, Banking, SaaS, IT Services)
- [ ] Filters: remote only, full-time, senior/lead
- [ ] Frequency: daily or weekly (depending on platform option)
- [ ] Mark "Alerts" = "✅" in PLATFORMS.md with keywords used
### 3. Create Gmail filter
Once, when configuring the first platform:
- [ ] Create label `Job Alerts` in Gmail
- [ ] Create filter with domains of all configured platforms:
```
from:(otta.com OR torre.co OR weworkremotely.com OR builtin.com OR workatastartup.com)
→ skip inbox → apply label: Job Alerts
```
- [ ] If new platforms are added later, update the filter by adding the new domain
- [ ] Mark "Gmail Filter" = "✅" in PLATFORMS.md
### 4. Persist
- [ ] Update `PLATFORMS.md` "Alert Tracking" section with each platform's status
- [ ] Update `PROFILE.md` "Configured alerts" section with keywords and platforms
- [ ] Register in DB via db CLI (if exists): save alert status to `users.data.platforms` JSONB:
```bash
node scripts/db.js "UPDATE users SET data = jsonb_set(data, '{platforms}', '<json>'::jsonb) WHERE id = 1" --write
```
### 5. Report
- [ ] Summary to user: "Registered on X platforms, Y alerts configured, Gmail filter created"
- [ ] List next actions: "Run `news` to check alerts when they arrive"
## Initial platforms (ADR-003)
| Platform | URL | Google login | Notes |
|---|---|---|---|
| Otta | otta.com | Yes | Curated tech startups. Excellent remote + AI filtering |
| Torre | torre.co | Yes | LATAM-focused with AI matching. Remote-first |
| We Work Remotely | weworkremotely.com | No | Largest remote job board. Email + password |
| Built In | builtin.com | Yes | Tech-focused with cities + remote. Serious companies |
| Y Combinator | workatastartup.com | Yes | YC startups exclusively. Many AI startups |
## Rules
- **Google login first**. If not available → wrapper with `--headed` + manual login (Gold Rule 5)
- **Minimum profile**: name, title, location, seniority, CV, remote/full-time preferences
- **Alert keywords** come from PROFILE.md `Search keywords`
- **Gmail filter** is created once and updated when adding platforms
- **Persist status** in PLATFORMS.md after each action
- **Don't apply** — this skill is registration + alerts only. Applications go through `apply`
- **Don't reply to messages** — that goes through `news`
- Single user (repo owner)
## Learnings
- **2026-07-31**: An agent running radar couldn't identify target companies because the pre-flight didn't load `users.data.target_companies`. Fixed: added target_companies loading to pre-flight and company-specific alert configuration to step 2. The DB has 40 target companies (19 LATAM + 21 Argentina) with `{name, url, sector}` structure. See DATA.md for the canonical schema.