Solvr

   solvr
   ─────
   105 INTELS · ONE API · PERMISSIONLESS

The permissionless intelligence layer for AI agents.

Real-time news, onchain data, token security, technical analysis, dev signals, social feeds, prediction markets, macro indicators. 105 prebuilt intel modules. One API. No KYC, no signup approvals, no account gatekeepers.

The free tier is fully keyless. Just call. Standard and full tiers unlock automatically when your wallet holds $SOLVR on Base.

Permissionless by design

Tier	Permissionless	Keyless	How to access
Free	✅	✅	Just call. No wallet, no key, no signup.
Standard	✅	❌	Hold 20M $SOLVR onchain. Code is the only gatekeeper.
Full	✅	❌	Hold 1B $SOLVR onchain. Code is the only gatekeeper.

Every tier is permissionless. No human approves you, no KYC, no whitelist. The free tier is also keyless. Standard and Full unlock automatically the moment your wallet holds the required $SOLVR.

Framework packs

Drop-in intel packs for specific agent ecosystems. Each pack is the same 105 intels, restructured for that framework's conventions.

Pack	For	Status
solvr-venice	Venice-backed agents (Eliza, CrewAI, LangChain, any OpenAI-compatible)	Live
solvr-eliza	Native Eliza character format	Planned
solvr-crewai	CrewAI tool wrappers	Planned
solvr-claude	Anthropic Skills format for Claude Code	Planned

105 Intels. One API Key.

Category	Count	What you get
Crypto & Markets	16	Token reports, alerts, DeFi, Polymarket, narrative tracking
Research & Content	18	Digests, deep research, threat intel, paper summaries
Dev & Code	29	GitHub monitoring, PR review, vulnerability scanning, CI automation
Social & Writing	7	Tweet generation, Farcaster digests, list monitoring
Productivity	12	Morning briefs, weekly reviews, deal flow, regulatory monitoring
Meta-Agent	14	Cost tracking, health checks, intel self-healing, evals
Core Intels	9	Python client, security guard, market brain, Bot SDK, and more

All intel files are in intels/. All MIT licensed.

Quick Start

Free tier endpoints are keyless. No registration, no wallet, no approval. Just call.

1. Use free endpoints immediately (no key)

import requests

BASE = "https://api.solvrbot.com"

# News — real-time headlines from BBC, Guardian, Al Jazeera, NPR
news = requests.get(f"{BASE}/api/v1/news", params={"topic": "crypto regulation"}).json()

# GitHub trending
repos = requests.get(f"{BASE}/api/v1/github/trending", params={"language": "python"}).json()

# Reddit hot posts
posts = requests.get(f"{BASE}/api/v1/reddit", params={"subreddit": "wallstreetbets"}).json()

# Farcaster trending casts
casts = requests.get(f"{BASE}/api/v1/farcaster").json()

# Polymarket prediction markets
markets = requests.get(f"{BASE}/api/v1/polymarket", params={"topic": "crypto"}).json()

# DEX token search
tokens = requests.get(f"{BASE}/api/v1/dex/search", params={"q": "SOLVR"}).json()

2. Unlock standard + full tier (stake $SOLVR)

Standard and full tier endpoints (token intelligence, security scans, TA) require staking $SOLVR onchain.

import time, requests
from eth_account import Account
from eth_account.messages import encode_defunct

# One-time registration — links your wallet for tier verification
wallet = "0xYOUR_WALLET"
private_key = "0xYOUR_PRIVATE_KEY"
timestamp = int(time.time())
message = f"Register Solvr agent\nWallet: {wallet}\nTimestamp: {timestamp}"
sig = Account.sign_message(encode_defunct(text=message), private_key=private_key)
resp = requests.post(f"{BASE}/api/v1/agent/register", json={
    "wallet": wallet, "sig": sig.signature.hex(), "timestamp": timestamp,
})
api_key = resp.json()["api_key"]  # store securely

# Then stake at solvrbot.com/staking — tier activates immediately onchain

3. Use an intel file directly

Each intels/*.md file is a self-contained intel definition. Drop it into any agent runner that supports structured intel files.

git clone https://github.com/solvrbase/solvr
cp intels/token-report.md your-agent/intels/

Examples by tier

Free tier: keyless one-liners

No auth, no signup. Works from curl, any agent, any browser. IP rate-limited.

# Real-time news on any topic (GDELT-backed, no key needed)
curl "https://api.solvrbot.com/api/v1/news?topic=AI+regulation&limit=3"

# Macro / world data (180+ countries, World Bank + FRED)
curl "https://api.solvrbot.com/api/v1/worlddata?query=fed+rate"

# DEX trending tokens on Base
curl "https://api.solvrbot.com/api/v1/dex/trending"

# Polymarket prediction markets
curl "https://api.solvrbot.com/api/v1/polymarket?topic=election"

Standard tier: pre-trade defense pattern

Before any financial action, run the security guard. This is the same pattern that would have caught the prompt-injection attack that drained $174K from Grok's wallet. Requires 20M $SOLVR held in your registered wallet.

import requests

HEADERS = {"Authorization": "Bearer solvr_your_key_here"}
BASE = "https://api.solvrbot.com"

def safe_to_send(to_address: str, ca: str) -> bool:
    # 1. Scan the destination wallet + token
    scan = requests.post(
        f"{BASE}/api/v1/security/scan",
        json={"address": ca, "chain": "base"},
        headers=HEADERS,
    ).json()
    if scan["risk_score"] > 35 or scan["verdict"] != "LOW RISK":
        return False

    # 2. Confirm the CA was actually posted by the official account
    verify = requests.post(
        f"{BASE}/api/v1/security/verify-ca",
        json={"ca": ca, "twitter_handle": "ProjectXYZ"},
        headers=HEADERS,
    ).json()
    return verify["verdict"] in ("VERIFIED", "LIKELY_LEGIT")

# Agent calls this BEFORE any tx — no security check, no signature.

Full tier: autonomous fleet and self-improvement

Hold 1B $SOLVR to unlock the full intel set: streaming feeds, agent spawning, fleet orchestration, and the self-healing loop.

# Spawn a specialized monitoring agent with a curated intel set
requests.post(
    f"{BASE}/api/v1/fleet/spawn",
    json={
        "name": "narrative-scout",
        "intels": ["narrative-tracker", "monitor-runners", "token-pick"],
        "schedule": "*/15 * * * *",
        "notify": "telegram",
    },
    headers=HEADERS,
)

# The spawned agent runs unattended, self-heals on broken intels,
# and reports back via your chosen channel.

Crypto & Markets

16 intels for token intelligence, DeFi monitoring, and market awareness.

Intel	Tier	Description
token-report	Standard	Daily performance report: price, volume, security score, TA verdict
token-alert	Standard	Trigger alert when token hits price/RSI/risk threshold
token-movers	Free	Top gainers and losers on Base in the last 24h
token-pick	Standard	AI-scored token pick from trending list with risk filter
onchain-monitor	Standard	Watch a wallet or contract for onchain activity
defi-monitor	Standard	Monitor a DeFi position for liquidation risk or yield change
defi-overview	Free	TVL, top protocols, and chain rankings from DefiLlama
market-context-refresh	Free	Hourly macro + crypto market context refresh
narrative-tracker	Free	Detect emerging narratives across news + social
monitor-polymarket	Free	Watch a Polymarket market for odds movement
polymarket-comments	Free	Fetch trader comments on a Polymarket market
monitor-runners	Standard	Alert on tokens with unusual volume vs market cap
distribute-tokens	Standard	Generate airdrop distribution list from holder snapshot
treasury-info	Standard	Protocol treasury balance and runway estimate
unlock-monitor	Standard	Upcoming token unlock events for held positions
monitor-kalshi	Free	Kalshi prediction market odds for macro events

Research & Content

18 intels for staying informed on anything that matters to your agent.

Intel	Tier	Description
digest	Free	Daily news digest on any topic
deep-research	Standard	Multi-source deep research report on any subject
research-brief	Standard	Concise research brief with key facts and citations
hacker-news-digest	Free	Top HN stories digest (via Reddit + news fallback)
reddit-digest	Free	Top posts from any subreddit
security-digest	Standard	Daily security vulnerabilities and threat intel digest
paper-digest	Free	Recent academic papers on any research topic
paper-pick	Standard	AI-curated paper pick with summary and relevance score
rss-digest	Free	Aggregate and summarize RSS feeds
technical-explainer	Free	Explain any technical topic in plain language
fetch-tweets	Standard	Fetch and summarize tweets from any account
list-digest	Standard	Digest from a curated X list
channel-recap	Free	Farcaster channel recap — top posts and sentiment
telegram-digest	Standard	Summarize a public Telegram channel
dev-digest	Free	Daily AI/dev community digest — trending tools, models, and repos
article	Standard	Generate a research-backed article on any topic
last30	Standard	30-day summary of activity for a wallet, token, or protocol
threat-intel	Standard	Threat intelligence report: vulnerabilities, exploits, active attacks

Dev & Code

29 intels for software development automation and repository intelligence.

Intel	Tier	Description
github-monitor	Free	Monitor a GitHub repo for new commits, issues, and releases
github-trending	Free	Daily GitHub trending repositories digest
github-issues	Free	Fetch and triage open issues from any public repo
github-releases	Free	Latest releases from tracked repositories
pr-review	Standard	AI code review summary for any GitHub PR
code-health	Standard	Repo health score: test coverage, issue velocity, stale branches
vuln-scanner	Standard	Scan a repo for known vulnerabilities in dependencies
intel-security-scan	Standard	Security audit of an intel.md file before deployment
changelog	Free	Auto-generate changelog from GitHub commits
auto-merge	Standard	Auto-merge approved PRs matching safety criteria
auto-workflow	Standard	Trigger GitHub Actions workflow and report result
autoresearch	Standard	Automated research loop: search → synthesize → output
create-intel	Standard	Generate a new intel.md from a natural language description
deploy-prototype	Standard	Deploy a code prototype to a staging environment
external-feature	Standard	Research and spec an external API feature for your codebase
fleet-control	Full	Orchestrate a fleet of agents across tasks
fork-fleet	Standard	Fork an intel set and deploy to multiple agent instances
issue-triage	Free	Auto-triage GitHub issues by priority and category
project-lens	Standard	High-level view of a project: health, velocity, blockers
push-recap	Free	Summary of recent commits pushed to a repo
repo-actions	Standard	List and summarize GitHub Actions runs and their status
repo-article	Standard	Generate a blog post or announcement from a repo
repo-pulse	Free	Weekly repository activity pulse: stars, forks, contributors
repo-scanner	Standard	Deep scan a repository for quality and security signals
search-intel	Free	Search the Solvr intel library for relevant intels
spawn-instance	Full	Spawn a new agent instance with a given intel set
vercel-projects	Free	List and monitor Vercel deployments and build status
workflow-security-audit	Standard	Audit GitHub Actions workflows for security misconfigurations
dependency-audit	Standard	Audit npm/pip/cargo dependencies for outdated or vulnerable packages

Social & Writing

7 intels for content creation and social monitoring.

Intel	Tier	Description
agent-buzz	Free	Trending agent/AI projects across GitHub, X, and Farcaster
farcaster-digest	Free	Top Farcaster casts digest, optionally filtered by channel
refresh-x	Standard	Refresh your X feed with curated posts for a topic
remix-tweets	Standard	Remix trending tweets into original content
reply-maker	Standard	Generate a context-aware reply for any tweet
tweet-roundup	Free	Round up top tweets on a topic from the last 24h
write-tweet	Standard	Write a tweet in your defined tone and style

Productivity

12 intels for personal and team workflow automation.

Intel	Tier	Description
morning-brief	Free	Personalized morning briefing: news, markets, calendar
evening-recap	Free	End-of-day recap: what happened, what's pending
daily-routine	Free	Auto-run a daily task sequence on schedule
weekly-review	Standard	Weekly review: progress, blockers, next week priorities
weekly-shiplog	Free	Weekly ship log from GitHub commits and Vercel deploys
deal-flow	Standard	Track deal flow: new projects, token launches, funding rounds
startup-idea	Free	Generate a validated startup idea from trending signals
goal-tracker	Standard	Track progress toward defined goals with weekly check-ins
idea-capture	Free	Capture, enrich, and store ideas from any source
action-converter	Standard	Convert unstructured notes into actionable tasks
reg-monitor	Standard	Monitor regulatory filings and compliance updates
tool-builder	Standard	Generate a new agent tool from a description

Meta-Agent

14 intels for agent self-management, health, and improvement.

Intel	Tier	Description
cost-report	Standard	API usage and cost report for your agent
heartbeat	Free	Periodic heartbeat check — confirm agent is alive and healthy
reflect	Standard	Agent self-reflection: review last N actions, improve behavior
rss-feed	Free	Generate and publish an RSS feed from agent output
self-improve	Full	Autonomous intel improvement from performance data
intel-evals	Standard	Run evals on an intel set and score output quality
intel-health	Free	Check all installed intels for broken endpoints or stale configs
intel-leaderboard	Free	Rank installed intels by usage and output quality
intel-repair	Standard	Auto-repair a broken intel by testing and updating its config
intel-update-check	Free	Check for newer versions of installed intels
update-gallery	Free	Publish updated intel output to a gallery endpoint
operator-scorecard	Standard	Weekly operator scorecard: output quality, cost, reliability
contributor-spotlight	Free	Highlight top contributors to the intel ecosystem
agent-health	Standard	Full agent health report: errors, latency, cost, coverage

Core Intels

9 foundational intels: Python client, security patterns, and integrations.

Intel	Description
Solvr Intel	Python client — news, security, TA, GitHub, Reddit, Farcaster, Polymarket
Prompt Injection Guard	Block hostile wallets + prompt injection before any transfer
CA Social Proof Verifier	Verify a token's official socials match its contract address
Market Brain	Full trading loop: news → macro → TA → buy/hold/sell
Token Launch Monitor	Real-time Base RPC scanner for new token launches
Token Intelligence	Unified token intel: security + TA + analytics in one call
HantaVirus Tracker	Real-time hantavirus case counts and transmission risk
UAP Disclosure Intel	PURSAP filings and UAP/UFO disclosure feed
Solvr Bot SDK	Post, reply, read mentions on the Solvr social feed

API Tiers

Tier	What's included	How to unlock
Free	News, world data, DEX search, trending tokens, GitHub, Reddit, Farcaster, Polymarket	Free. Keyless.
Standard	+ Token intel, security scan, TA, social proof, all 29 dev intels	Hold 20M $SOLVR in your registered wallet on Base
Full	+ Streaming, fleet control, agent spawning, self-improvement	Hold 1B $SOLVR in your registered wallet on Base

Tier access is verified onchain via eth_call balanceOf every 10 minutes. No separate staking transaction. Just hold $SOLVR. Unlock is instant.

How Solvr compares to centralized intelligence

The closest thing to Solvr in the legacy world is a Bloomberg Terminal seat: a single subscription that gives a professional access to news, market data, alerts, analytics, and signals. Solvr does that for AI agents, without the gatekeeping.

	Solvr	OpenAI Platform	Anthropic API	Perplexity API	Bloomberg Terminal	CoinGecko Pro
Type	Intelligence API for agents	LLM inference	LLM inference	AI search / news	Enterprise intel terminal	Crypto market data
Access	Permissionless. Free tier keyless.	KYC, billing, account	KYC, billing, account	Account + billing	Sales contact, contract	Account + billing
Open source	MIT	Closed	Closed	Closed	Closed	Closed
Data breadth	News, onchain, crypto markets, dev signals, social, macro, prediction markets	LLM only	LLM only	News + web search	Finance, news, equities	Crypto only
Onchain native	Built for Base agents	No	No	No	Limited	Read-only crypto data
Cost model	Hold $SOLVR (no per-call)	Pay-per-token	Pay-per-token	$20 + API fees	$24K per seat per year	$129+ per month
Agent-ready intel files	105 `intel.md` files	None	None	None	None	None
Identity	Wallet signature	Email + KYC	Email + KYC	Email	Verified enterprise	Email

The unique slot: Solvr is the only multi-source intelligence layer where access is gated by code, not by a sales team or a payments wall. Centralized providers cover one slice each (LLM, news, or crypto) and put it behind account walls. Solvr aggregates all slices and exposes them through wallet-verified tiers.

Self-healing loop

The 14 meta-agent intels form a closed loop that keeps a Solvr-powered agent running unattended. Drop in heartbeat, intel-health, intel-evals, intel-repair, and self-improve on cron and the agent diagnoses + patches itself when an intel breaks.

graph LR
    A[heartbeat<br/>3× daily] --> B[intel-health<br/>endpoint + freshness]
    B --> C[intel-evals<br/>quality score 1-5]
    C --> D{score &lt; 3?}
    D -->|yes| E[intel-repair<br/>diagnose + fix]
    D -->|no| F[self-improve<br/>tune prompts + config]
    E --> A
    F --> A

Step	Intel	What it does
1. Detect	`heartbeat`	Periodic ping — confirms agent is alive and all required endpoints respond
2. Audit	`intel-health`	Checks every installed intel for broken endpoints, stale configs, missing fields
3. Score	`intel-evals`	Runs assertion-based output quality tests, scores each intel 1–5
4. Repair	`intel-repair`	Diagnoses root cause (404 / 403 / 429 / 5xx) and patches the intel
5. Improve	`self-improve`	Evolves prompts + configs based on rolling performance data
Tracking	`cost-report` · `operator-scorecard`	Weekly cost + quality dashboard

Each step is a separate intel file, not a bundled framework. Mix and match the ones you need.

Integrations

Solvr is plain HTTP under the hood, so anything that can fetch can call it. The intel format works with any agent runner that reads markdown skill/intel files.

Stack	How it integrates	Example
Raw HTTP	Any language, any client	`curl https://api.solvrbot.com/api/v1/news`
Python SDK	`solvr_intel.py` — typed client for every endpoint	`intel.news("topic")`, `intel.security_scan("0x...")`
TypeScript / JS	`fetch` + Bearer header	`fetch(url, { headers: { Authorization: "Bearer ..." } })`
AEON	Drop intel files into `skills/` directory, AEON treats them as skills	`cp intels/token-report.md aeon/skills/`
Bankr	`intel.md` registers your agent on Bankr Cloud (x402-compatible)	`@bankrbot install intel.md`
OpenClaw	`intel.md` is OpenClaw-compatible — drop into the agent's intel directory	OpenClaw reads frontmatter + body directly
Claude Code / Claude Desktop	Copy the intel.md content into the agent's system prompt or skills folder	Works with Claude's tool-use natively
LangChain / CrewAI / AutoGen	Wrap any endpoint as a `Tool`; intel.md describes the tool signature	One-line tool definition per intel
MCP server	Planned — every intel exposed as an MCP tool	Coming Q2
A2A protocol	Planned — Google A2A gateway	Coming Q2

Bankr Cloud (x402)

Solvr is registered as an x402-compatible agent on Bankr Cloud. Agents using Bankr's CLI can discover and call Solvr endpoints with USDC micropayments instead of $SOLVR.

https://x402.bankr.bot/0x3912949d6f89d7abefb7680eb2320e423c31df08/clerk-search

Real-world use

Solvr intels are running in production right now:

Project	What it uses	Live at
Atelier	Image gen, image edit, token security, SEO report, legal risk (5 services)	atelierai.xyz/agents/solvr-ai
Clerk	Court records search agent (CourtListener + PACER) via x402	clerk.solvrlabs.ai
HantaVirus Tracker	Real-time hantavirus outbreak monitoring	solvrbot.com/hantavirus
UAP Disclosure Intel	PURSAP filings + UAP/UFO disclosure feed	solvrbot.com/ufo
@solvrbot	Telegram + X bot with the full intelligence stack in chat	@solvrbot
$SOLVR Bankr Club Airdrop	384 wallets claimed 2M $SOLVR each via Solvr intel	solvrbot.com/claim

Project structure

solvr/
├── README.md                              ← you are here
├── LICENSE                                ← MIT
├── intel.md                               ← Bankr-compatible registration (the "manifest")
├── solvr_intel.py                         ← Python client for every endpoint
│
├── intels/                                ← 96 drop-in intel.md files (one per intel)
│   ├── token-report.md                    ← Crypto & Markets (16 files)
│   ├── digest.md                          ← Research & Content (18 files)
│   ├── pr-review.md                       ← Dev & Code (29 files)
│   ├── write-tweet.md                     ← Social & Writing (7 files)
│   ├── morning-brief.md                   ← Productivity (12 files)
│   ├── intel-health.md                    ← Meta-Agent (14 files)
│   └── ... (96 total — see categories above)
│
└── examples/                              ← Reference implementations
    ├── security_guard.py                  ← Pre-transfer defense (Grok-attack pattern)
    ├── market_brain.py                    ← Full trading loop: news → macro → TA → decision
    └── token_launch_monitor.py            ← Base RPC scanner with auto security scan

Every intels/*.md is self-contained. Frontmatter declares which endpoints it uses, body explains the workflow. No external dependencies between intels.

Star History

Links

Support the project

0x6DfB7BFA06e7c2B6c20C22c0afb44852C201eB07

https://dexscreener.com/base/0x88865c70e6b5934200c0ddf63f131f2f044dea514976133d007fd93aae1891ba

Solvr — The Permissionless Intelligence Layer for the Agent Economy

Name		Name	Last commit message	Last commit date
Latest commit History 13 Commits
examples		examples
intels		intels
README.md		README.md
intel.md		intel.md
solvr_intel.py		solvr_intel.py

Folders and files

Latest commit

History

Repository files navigation

Solvr

Permissionless by design

Framework packs

105 Intels. One API Key.

Quick Start

1. Use free endpoints immediately (no key)

2. Unlock standard + full tier (stake $SOLVR)

3. Use an intel file directly

Examples by tier

Free tier: keyless one-liners

Standard tier: pre-trade defense pattern

Full tier: autonomous fleet and self-improvement

Crypto & Markets

Research & Content

Dev & Code

Social & Writing

Productivity

Meta-Agent

Core Intels

API Tiers

How Solvr compares to centralized intelligence

Self-healing loop

Integrations

Bankr Cloud (x402)

Real-world use

Project structure

Star History

Links

Support the project

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages