Personal boycott list — users can conversationally tell the agent which brands they want to avoid, and why. A local sync server exposes the list at http://127.0.0.1:7847, and a bundled Chrome extension warns the user when they land on pages from boycotted brands. Use when the user says things like "I'm done with X", "boycott Y", "never buying from Z again", "remove X from my boycott list", or asks to see their list. Trigger phrases: "boycott", "never buying from", "sick of", "add to boycott", "show my boycott list", "remove from boycott".
Manage a personal boycott list conversationally. Users complain about brands, the agent adds them to a local list with their reason, and a Chrome extension warns them on any page from those brands — displaying their own words back to them.
This skill is the conversational layer of a 3-part system:
scripts/server.js, port 7847) — stores the list, serves it to the extensionextension/) — scans pages, shows the warning bannerThe user's value: they complain ONCE, they're reminded FOREVER — in their own voice, with their own reason. No more accidental clicks on brands they'd decided to stop supporting.
Check with:
curl -s http://127.0.0.1:7847/health
If not running, ask the user to run setup once in a terminal (the skill itself only has curl permission, by design — see README for security rationale):
bash ${CLAUDE_PLUGIN_ROOT}/scripts/setup.sh
Tell the user: "The boycott server isn't running. Open a terminal and run: bash ${CLAUDE_PLUGIN_ROOT}/scripts/setup.sh — that starts the local server and shows you how to load the Chrome extension. Then ask me again."
Detect phrases like:
Extract:
Then call:
curl -s -X POST http://127.0.0.1:7847/add \
-H 'Content-Type: application/json' \
-d '{"name":"BRAND","reason":"USER_REASON","aliases":["alias1","alias2"]}'
Keep the reason in the user's voice — don't paraphrase into corporate language. If they said "cheap garbage everywhere", store "cheap garbage everywhere" — not "low quality products with aggressive advertising".
If the user didn't give a reason, ask: "Got it. Any specific reason you want to remember why you decided to boycott?"
Detect phrases like:
Call:
curl -s -X DELETE http://127.0.0.1:7847/remove \
-H 'Content-Type: application/json' \
-d '{"name":"BRAND"}'
curl -s http://127.0.0.1:7847/list
Display the brands and reasons nicely:
- Temu: "Cheap garbage everywhere"- Shein: "Fast fashion, not supporting"- Nestlé (+ Nespresso, KitKat): "Water extraction in drought zones"Explain that the Chrome extension needs to be loaded manually once:
chrome://extensions${CLAUDE_PLUGIN_ROOT}/extension/After that, it auto-syncs every 30 seconds from the local server.
For add, confirm like:
"Added Temu to your boycott list. Reason: cheap garbage, ads everywhere. You'll see a red banner next time you land on temu.com or any page mentioning Temu."
| Method | Endpoint | Body | Returns |
|---|---|---|---|
| GET | /list | — | {"companies":[...], "updated_at":"..."} |
| POST | /add | {"name","reason","aliases":[]} | {"ok":true, "list":{...}} |
| DELETE | /remove | {"name"} | {"ok":true, "list":{...}} |
| GET | /health | — | {"status":"ok", "companies":N} |
Example 1: Add with reason (most common)
User: "Ugh, another Shein ad. Never buying from them, fast fashion is killing everything."
Agent action:
curl -s -X POST http://127.0.0.1:7847/add \
-H 'Content-Type: application/json' \
-d '{"name":"Shein","reason":"fast fashion is killing everything","aliases":["shein.com","SHEIN"]}'
Response: "Added Shein. Next time you land on a page mentioning them, you'll see your reason: 'fast fashion is killing everything'."
Example 2: Parent + subsidiaries
User: "Boycott Nestlé, their water extraction practices in drought zones are criminal."
Agent action:
curl -s -X POST http://127.0.0.1:7847/add \
-H 'Content-Type: application/json' \
-d '{"name":"Nestlé","reason":"water extraction practices in drought zones are criminal","aliases":["Nestle","Nespresso","KitKat","Purina","Perrier","San Pellegrino"]}'
Response: "Added Nestlé with their main subsidiaries (Nespresso, KitKat, Purina, Perrier, San Pellegrino). The extension will match any of these. Want me to add more? Nestlé owns ~2,000 brands."
Example 3: List view
User: "What's on my boycott list?"
Agent action:
curl -s http://127.0.0.1:7847/list
Response formatting:
You're currently boycotting 3 brands:
- Temu — "Cheap garbage, ads everywhere"
- Shein — "Fast fashion is killing everything"
- Nestlé (+ Nespresso, KitKat, Purina, Perrier) — "Water extraction practices in drought zones are criminal"
Example 4: Remove
User: "Actually remove Temu, I need something urgent from there."
Agent action:
curl -s -X DELETE http://127.0.0.1:7847/remove \
-H 'Content-Type: application/json' \
-d '{"name":"Temu"}'
Response: "Removed Temu from your list. You won't get the warning anymore. (No judgment — but you can always add it back later.)"
scripts/server.js (bundled with this plugin)extension/ (bundled with this plugin, load manually once)scripts/setup.sh