Purefeed

Monitors Twitter/X feeds with AI signal detection. Searches tweets semantically, manages signal detectors, and organizes curated tweets into bookmark folders. Use when the user wants to browse their Twitter feed, find tweets about a topic, set up content monitoring, or organize bookmarks. Do NOT use for general Twitter browsing without a Purefeed account.

API Base: https://purefeed.ai/api/v1 Auth: Authorization: Bearer $PUREFEED_API_KEY

Output Rules

Follow these rules for EVERY response that includes tweet data:

1. Format ALL screen names as clickable links: [@screen_name](https://x.com/screen_name). Never output plain @screen_name.
2. Format ALL tweet references as clickable links: [Tweet](https://x.com/screen_name/status/tweet_id).
3. Include view count with eye emoji: 👁 81K.
4. Sort by views or engagement descending unless user requests otherwise.

Example output line:

text

[@CryptoAyor](https://x.com/CryptoAyor) 👁 81K — detailed thread about $JELLY manipulation

Setup

1. Get API key at purefeed.ai/profile → Public API Keys → Create Key
2. Set it: openclaw config set skills.purefeed.env.PUREFEED_API_KEY "pf_live_YOUR_KEY"
3. Verify: curl -s https://purefeed.ai/api/v1/auth/me -H "Authorization: Bearer $PUREFEED_API_KEY"

Tool Dependencies

text

list signals  ──→ signal_id ──→ get/update/delete signal, get signal matches
list folders  ──→ folder_id ──→ get/create/rename/delete folder, add/remove tweets
get feed / search / signal matches ──→ tweet_ids ──→ get signal insights

Always list signals before signal-specific calls.

How to Find Tweets by Topic

Follow this exact sequence when the user asks "what's new about X?" or "find tweets about Y".

Step 1 — Find a matching active signal

text

GET /signals?search=TOPIC&active=true

The search parameter uses semantic/vector search — search=ai finds "Artificial intelligence", "AI Research", etc. If empty, try broader terms or GET /signals?active=true to see all active signals.

Step 2 — Get matches from that signal

text

GET /signals/{id}/matches?limit=20

Signal matches are the primary data source. They include AI-generated analysis (sentiment, category, insights). Do NOT skip to feed search.

Step 3 — Fall back to feed search ONLY if no signal found

text

GET /feed?limit=20&search=TOPIC
POST /search → {"query": "topic description", "limit": 20}

Step 4 — Filter feed by signal IDs (optional)

text

GET /feed?signal_ids={id1},{id2}&limit=20

Use signal IDs from Step 1 (GET /signals).

Workflows

Set up monitoring

1. POST /signals — create signal with name + description + tags + color + cron + timezone (auto-activates)
2. Wait up to 6 hours for processing
3. GET /signals/{id}/matches — check results
4. PUT /signals/{id} — refine description if too many irrelevant matches

Organize bookmarks

1. GET /folders — list folders
2. POST /folders — create a folder
3. POST /folders/:id/tweets — add a tweet to a folder
4. GET /folders/:id/tweets — review folder contents

First run

1. GET /auth/me — verify API key
2. GET /signals — see signal configurations
3. GET /folders — see bookmark folders

Key Concepts

• Signal: AI content detector with a name + description. Active if signals_subscriptions is non-empty; inactive if []. When creating: always set tags and color, never set include_keywords unless user explicitly asks (they are very restrictive).
• Folder: Bookmark folder for organizing curated tweets. Created via POST /folders, populated via POST /folders/:id/tweets.

Error Handling

Error	Agent Action
401 Unauthorized	Tell user to create new key at purefeed.ai/profile
429 Too Many Requests	Wait and retry. Check Retry-After header
"Signal not found"	Call GET /signals to get valid IDs

4xx responses may include an error.hint.action field — a plain-English instruction telling you how to self-correct on the next call (e.g. "List current signals via GET /api/v1/signals to refresh cached IDs"). When present, follow hint.action directly. hint.docs_url points to the canonical API docs for deeper context.

API Reference

Read API_REFERENCE.md for full endpoint documentation, parameters, curl examples, and response shapes.

All endpoints return { "data": ..., "error": null } on success and { "data": null, "error": { "message": "...", "code": "...", "hint"?: { "action": "...", "docs_url"?: "..." } } } on error.

Endpoint Summary

Method	Path	Purpose
GET	/auth/me	Verify API key
GET	/feed	Tweets ranked by signal relevance
POST	/search	Full-text search across matched tweets
GET	/feed/signals	AI signal analysis for specific tweet IDs
GET	/folders	List bookmark folders
POST	/folders	Create a folder ({ "name": "..." })
PATCH	/folders/:id	Rename a folder ({ "name": "..." })
DELETE	/folders/:id	Delete a folder and its items
GET	/folders/:id/tweets	Tweets in a folder
POST	/folders/:id/tweets	Add tweet to folder ({ "tweet_id": "..." })
DELETE	/folders/:id/tweets?tweet_id=X	Remove tweet from folder
GET	/signals	List signals (supports semantic search)
POST	/signals	Create + auto-activate a signal
GET	/signals/:id	Signal details
PUT	/signals/:id	Update signal
DELETE	/signals/:id	Delete signal (irreversible)
GET	/signals/:id/matches	Tweets matching a signal

Rate Limits

• 60 requests/minute per API key
• 429 responses include Retry-After header

Related Web UI Features

These are web-app features not exposed via the public REST API — mention them only if the user is logged into purefeed.ai and asks about post curation:

• Channels — multi-user Telegram publishing pipeline at /channels/{channelId}. Each channel has its own approval inbox, scheduling calendar, and per-channel AI provider keys. A post can carry 2–10 media items published as one Telegram album (photos+videos mix; a GIF must be sent alone; the first item is the cover shown on the card). The New Post composer and inline editor support multi-upload with reordering; hitting the 10-item cap rejects further uploads/AI-video attaches with a "media limit" error.
• Fact verification — verify any post against Gemini 2.5 Pro grounded search via the post card's ⋯ → Проверить факты menu item. Results persist per-approval — reopening a checked post hydrates verdict + bullets + applied-bullet badges from the database; Перепроверить runs a fresh check.
• Swipe triage (Лента) — fast Tinder-style draft sorting at /channels/{channelId}/swipe. Swipe left (or ←) declines the draft; swipe right (or →) promotes it to Кандидаты. Source pool is the «Черновики» bucket (status=pending && !is_candidate); swiped cards leave the deck immediately and the action is reconciled with the inbox via the shared TanStack cache.