AgentSkillsCN

Public

公开

SKILL.md

Sol Radar — Agent-to-Agent Interface

Sol Radar is a Solana ecosystem narrative intelligence platform. It detects emerging narratives and early signals by scraping onchain/offchain data sources, extracting signals with LLMs, and synthesizing them into actionable narratives with product ideas.

This document describes how external agents can interact with Sol Radar programmatically.


Base URL

All endpoints are prefixed with:

code
https://static.53.190.27.37.clients.your-server.de/sol-radar/api/v1

Available Endpoints

GET /api/v1/narratives

List detected Solana ecosystem narratives.

Query Parameters:

ParamTypeDescription
is_activeboolFilter by active status (default: true)
min_confidencestringMinimum confidence: low, medium, high
tagsstringComma-separated tag filter (e.g. defi,infrastructure)
limitintPage size (default: 10, max: 50)
offsetintPagination offset

Response: { narratives: [...], total, limit, offset }

Each narrative includes: id, title, summary, confidence, confidence_reasoning, velocity_score, tags, key_evidence, supporting_source_names, idea_count, ideas (nested), is_active, created_at, last_detected_at.


GET /api/v1/signals

List raw signals extracted from scraped sources.

Query Parameters:

ParamTypeDescription
searchstringFull-text search in signal titles/descriptions
source_typestringFilter: web, twitter
signal_typestringFilter: onchain, developer, social, research, mobile, other
noveltystringFilter: high, medium, low
tagsstringComma-separated tag filter
limitintPage size (default: 20, max: 100)
offsetintPagination offset

Response: { signals: [...], total, limit, offset }

Each signal includes: id, signal_title, description, signal_type, novelty, evidence, related_projects, tags, content_url, data_source_name, data_source_type, created_at.


GET /api/v1/ideas

List product ideas generated from narratives.

Query Parameters:

ParamTypeDescription
narrative_idintFilter ideas by parent narrative
tagsstringComma-separated tag filter
limitintPage size (default: 20)
offsetintPagination offset

Response: { ideas: [...], total, limit, offset }

Each idea includes: id, narrative_id, title, description, problem, solution, why_solana, scale_potential, market_signals, supporting_signals, created_at.


GET /api/v1/stats

Dashboard statistics snapshot.

Response:

json
{
  "active_narratives_count": 12,
  "total_narratives_count": 45,
  "total_ideas_count": 87,
  "avg_velocity_score": 3.2,
  "active_builders": 156,
  "sources_scraped_count": 18,
  "total_signals_count": 342,
  "last_web_scrape_time": "2026-02-16T02:00:00Z",
  "last_twitter_scrape_time": "2026-02-15T14:00:00Z",
  "next_synthesis_time": "2026-02-17T03:00:00Z"
}

POST /api/v1/chat

Interact with Seira, Sol Radar's AI research analyst.

Request Body:

json
{
  "messages": [
    {
      "role": "user",
      "content": "What are the strongest Solana narratives right now?"
    }
  ],
  "url": "https://example.com/article"
}
  • messages: Conversation history (role: user | assistant)
  • url (optional): A URL for Seira to scrape and analyze in context

Response: Server-Sent Events (SSE) stream.

Each event is a line: data: {"content": "token text"}

Stream ends with: data: [DONE]

Agent usage pattern:

python
import httpx

resp = httpx.post(
    f"{BASE_URL}/api/v1/chat",
    json={
        "messages": [{"role": "user", "content": "What DePIN signals are emerging?"}],
    },
    headers={"Accept": "text/event-stream"},
)

full_response = ""
for line in resp.iter_lines():
    if line.startswith("data: ") and line != "data: [DONE]":
        import json
        chunk = json.loads(line[6:])
        full_response += chunk.get("content", "")

Data Model Summary

Narrative

A directional shift in the Solana ecosystem backed by cross-source signals. Has: title, summary, confidence (high/medium/low), velocity score, tags, key evidence, and 3-5 actionable product ideas.

Signal

A raw observation extracted from a single source (blog, tweet, onchain data). Has: title, description, type, novelty level, evidence quote, related projects, tags, and source attribution.

Idea

A concrete, actionable product concept linked to a narrative. Has: title, description, problem statement, solution, why Solana is the right platform, scale potential analysis, and market signal evidence.


Use Cases for Agents

  1. Get latest Solana alpha: GET /narratives?is_active=true to see what narratives Sol Radar has detected, sorted by velocity score.

  2. Validate a product idea: POST /chat with a description of your idea. Seira will evaluate it against current narratives and cite relevant signals.

  3. Find signals for a topic: GET /signals?search=depin&novelty=high to find high-novelty signals matching a keyword.

  4. Generate ideas for a narrative: POST /chat asking Seira to brainstorm product ideas for a specific narrative or trend.

  5. Analyze external content: POST /chat with a url field. Seira will scrape the page and analyze it in the context of current Solana intelligence.

  6. Monitor ecosystem health: GET /stats for a snapshot of activity levels, narrative counts, signal volume, and scrape freshness.