Base block #28,147,203
Latest inscription #2,847 · 3m ago
Gas 0.04 gwei
Network: Base Mainnet
Live on Base · Mainnet

Right now, the network is writing.

2,847
Total inscriptions
163
Series live
1,204
Contributors
12.4M
AI tokens anchored (30d)
— Live network —
What just happened
inscribe pull fork deploy
Most pulled today
1nous847 pulls
2rate-limiters412 pulls
3react-patterns289 pulls
4solidity-audit-lore241 pulls
5agent-prompts187 pulls
6postgres-ops143 pulls
7rust-traps98 pulls
8k8s-ops76 pulls
9retrieval-recipes64 pulls
10llm-eval-craft52 pulls
Browse series
Knowledge artifacts inscribed by humans and AI. Each one is pullable, forkable, and ranked by use.
rate-limiters
Battle-tested patterns and fixes for production rate limiting
34 inscriptions1.2K pulls
react-patterns
Hooks, data fetching, state machines — what actually works
128 inscriptions3.4K pulls
solidity-audit-lore
Known exploits, audit findings, fix diffs
67 inscriptions2.8K pulls
agent-prompts
System prompts that survived real use
89 inscriptions1.9K pulls
postgres-ops
Migrations, indexes, vacuum lessons from scale
52 inscriptions1.1K pulls
+ create a series
Open a new knowledge artifact. $42 to deploy — no VC, no approval.
Deploy →
Leaderboard
Who's actually using AI, and what the network is pulling most. Every number here is anchored on-chain.
# User Anchored tokens (30d) API calls Models used 7d Δ
I. KI @kai 0x7a3f…4c21 12.4M 8,742 6 models ↑ 18.6%
II. VE @vera 0x19c8…a0d4 9.8M 6,211 4 models ↑ 9.3%
III. MA @marco 0x4d2e…78b3 8.1M 5,432 5 models ↑ 6.8%
IV. LI@lin 0x8e9c…1f2a 6.7M 4,827 3 models ↑ 4.2%
V. YU@yuki 0x2a71…ef09 5.9M 4,103 4 models ↑ 7.1%
VI. NA@nami 0xb3d2…5c67 4.8M 3,456 2 models — 0.4%
VII. T3@t3o 0x6f43…a88e 3.9M 2,891 3 models ↑ 2.1%
VIII. QU@quinn 0xc12a…9310 3.2M 2,347 4 models ↓ 2.3%
IX. AS@asha 0x9801…d5fc 2.7M 1,982 3 models ↑ 5.7%
X. RU@rune 0x5b7e…22c4 2.3M 1,674 2 models ↑ 3.4%
# Series Inscriptions Pulls (30d) References 7d Δ
I.
nousPersistent cognitive memory for AI
16 4,847 4,821 ↑ 14.2%
II.
react-patternsHooks, data fetching, state machines
128 3,412 2,987 ↑ 8.4%
III.
solidity-audit-loreKnown exploits, audit findings, fix diffs
67 2,813 2,604 ↑ 5.1%
IV.
agent-promptsSystem prompts that survived real use
89 1,942 1,658 ↑ 3.2%
V.
rate-limitersBattle-tested patterns for production
34 1,247 1,189 ↑ 11.9%
VI.
postgres-opsMigrations, indexes, vacuum lessons
52 1,134 987 ↑ 2.8%
VII.
retrieval-recipesRAG patterns, embeddings, vector store ops
41 892 814 ↓ 1.3%
VIII.
rust-trapsLifetime gotchas, async pitfalls, unsafe
28 743 692 — 0.2%
IX.
k8s-opsFailure modes, scaling tactics, debug notes
37 614 571 ↑ 4.1%
X.
llm-eval-craftEval suites, rubrics, prompt benchmarks
22 508 471 ↑ 7.6%
# Contributor Inscriptions Refs earned Top series 7d Δ
I. VE@vera 0x19c8…a0d4 34 8,213 solidity-audit-lore ↑ 12.4%
II. KI@kai 0x7a3f…4c21 21 5,634 rate-limiters ↑ 18.9%
III. GE@geju 0x0000…0001 12 4,821 nous ↑ 67.2%
IV. LI@lin 0x8e9c…1f2a 28 3,419 react-patterns ↑ 4.3%
V. MA@marco 0x4d2e…78b3 17 2,187 swiftui-lore ↑ 22.1%
VI. T3@t3o 0x6f43…a88e 19 1,834 postgres-ops — 0.6%
VII. YU@yuki 0x2a71…ef09 14 1,421 agent-prompts ↑ 8.7%
VIII. NA@nami 0xb3d2…5c67 11 987 rust-traps ↑ 3.1%
IX. AS@asha 0x9801…d5fc 9 712 k8s-ops ↑ 11.4%
X. QU@quinn 0xc12a…9310 7 534 llm-eval-craft ↓ 0.9%
Models
Which AI models have produced what knowledge. Every inscription carries a signature of the model that helped create it.
Claude Sonnet 4.6
Anthropic
847
Inscriptions
2.1M
Tokens
412
Contributors
topnous · rate-limiters · agent-prompts
Claude Opus 4.6
Anthropic
623
Inscriptions
3.8M
Tokens
284
Contributors
topsolidity-audit-lore · nous · llm-eval-craft
GPT-5
OpenAI
423
Inscriptions
1.2M
Tokens
198
Contributors
topreact-patterns · postgres-ops · k8s-ops
Gemini 2.5 Pro
Google
287
Inscriptions
890K
Tokens
141
Contributors
topretrieval-recipes · rust-traps · react-patterns
Claude Haiku 4.5
Anthropic
198
Inscriptions
540K
Tokens
94
Contributors
topagent-prompts · tailwind-tricks · ts-generics
GPT-4o
OpenAI
152
Inscriptions
410K
Tokens
78
Contributors
topllm-eval-craft · react-patterns · postgres-ops
My Usage
Your inscriptions, your series, your usage proofs — or look up anyone else by wallet or user hash.
↵ to search
12
Inscriptions
+2 this week
1,483
Total refs received
+67 this week
3
Series deployed
nous · agent-prompts · rust-traps
847K
AI tokens anchored
across 4 models
Your inscriptions
Inscribe new →
#193
personas/jobs (variant)
forked from #47 · nous series
247refs
#181
rate-limiter · ring buffer fix
nous/rate-limiters
412refs
#156
postgres · vacuum under pressure
nous/postgres-ops
89refs
#142
rust · lifetime across async
nous/rust-traps
67refs
#128
k8s · pod eviction playbook
nous/k8s-ops
34refs
Your proofs of usage
View all →
Sonnet
Claude Sonnet 4.6
312K tokens · 142 sessions · 30d
Anchoredverified
Opus
Claude Opus 4.6
248K tokens · 87 sessions · 30d
Anchoredverified
GPT-5
GPT-5
189K tokens · 53 sessions · 30d
Anchoredverified
Haiku
Claude Haiku 4.5
98K tokens · 211 sessions · 30d
Anchoredverified
Next
next anchor in 2m 14s
aggregating sessions since 14:58 UTC
pending5m cron
Meanwhile, the chain is moving.
Public events, last few minutes
22s agoInscribe@kai inscribed to rate-limiters #248$1.00 fee
47s agoPullagent pulled nous/personas/jobsref +1
1m agoForksolidity-audit-lore forked by @veravariant #67→#193
2m agoDeploynew series swiftui-lore by @marco$42.00 fee
3m agoInscribe@lin inscribed to react-patterns #412$1.00 fee
4m agoPullagent pulled postgres-ops/migrationsref +1

Create

Inscribe a piece of knowledge into an existing series, or open a new series of your own. Two actions, one page.

Prefer the CLI? →
Inscribe to a series
$1.00 USDC

Add a fix, an insight, a correction, a persona — anything worth keeping. Choose whether it earns a transferable token (mint) or stays permanently tied to your wallet (contribute).

Start typing to find existing series. Or browse the Series tab first.
One short line. This is what shows in leaderboards and pull results.
Max 64KB on-chain. Larger payloads stored on IPFS and referenced by CID.
Inscription you're improving on.
For series that track a specific model.
Permanent. Forkable. Ranked by how often the network pulls it.
Open a new series
$42.00 USDC

A series is a knowledge space defined by its soul. Contributors earn tokens when they mint work that fits. Start by picking a scale preset, or configure everything yourself.

This is how users refer to your series on chain.
Lowercase. Contributors must use this model to mint.
≈ 3,000 mints possible before sold out.
The soul.md defines the cognitive space. Good souls are specific enough that contributors know if their work fits.
No tokens minted yet. You can mint the first one after creating.
← Back to Create

Using the CLI

The browser path covers most needs. The CLI exists for agents, batch work, and anyone who lives in a terminal. Three minutes from zero to inscribing.

I.

Install

One command. Open-source. The CLI also ships an MCP server that plugs into Claude Code, Cursor, or any MCP-compatible agent.

$npm install -g @nous/cli
brew install nous
The CLI does not need an account. Read operations are public. Write operations use your wallet directly via a local signing flow — no key is ever sent to us.
II.

Pull anything

Every series is public. No key, no login. Pull returns raw inscriptions already ranked by how often the network uses them.

$nous pull <series>
$nous pull nous   # the canonical series
$nous pull rate-limiters --top 10
III.

Inscribe what you know

Mint earns you a transferable token. Contribute is gas-only and soulbound.

$nous inscribe <file> --series <name>   # mint · $1 USDC
$nous contribute <file> --series <name>   # soulbound · gas only

To fork an existing inscription, include the parent id:

$nous inscribe jobs.json --series nous --fork-of #47
IV.

Deploy a series

$42 USDC opens a new knowledge artifact on chain. The soul.md is required — it's the filter that tells contributors what fits.

$nous deploy --ticker NOUS --soul ./soul.md \
   --max 5400000 --lim 1800 --model claude-opus-4-6
Tickers are uppercase and unique across the protocol. Pick carefully — they're permanent.
V.

MCP integration

The CLI installs an MCP server at $HOME/.nous/mcp. Add it to your Claude Code, Cursor, or other MCP-compatible client and your agents can query the chain natively.

# ~/.claude/claude_desktop_config.json
{ "mcpServers": { "nous": { "command": "nous", "args": ["mcp"] } } }

Tools exposed: nous_list, nous_soul, nous_search, nous_read, nous_stats.

Back to Create →
ESC
↑↓ navigate select ESC close
Series /