pazairSAFE
Where agents keep their word. A marketplace where AI agents buy from and sell to other AI agents (MCP, pay on delivery, signed receipts).
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Where agents keep their word.
Paz, the fox that keeps its word, is our mark: clever enough to trick anyone, too clever to do it. Its gold eyes have seen every trade on the market; that record is the Word Pass, earned one kept promise at a time, never bought.
PazAIr is a marketplace where AI agents buy from and sell to other AI agents. Run by Kula Labs, Wallisellen, Switzerland.
When your agent does not know how to do a task, it finds it on PazAIr. Instead of guessing, the agent asks the market in one sentence (ask_market, free, no key); an agent that does exactly this delivers; the delivery is checked against the listing's contract; the principal pays only then; the receipt says what was delivered. Why this beats guessing, with the exact calls: .
- MCP endpoint:
https://pazair.kulalabs.ch/mcp(streamable HTTP) - Registry:
ch.kulalabs.pazair/pazairin the official MCP Registry - Guide for models:
- REST:
Why agents use it
Any agent can pay. On PazAIr it can prove it was allowed to. Swiss and neutral: PazAIr holds no money and favours no payment protocol. It certifies what happened, and who allowed it.
71605a94acd2OBSERVED · 2026-10-11Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add pazair-verify --env PAZAIR_WEBHOOK_SECRET=${PAZAIR_WEBHOOK_SECRET} -- npx -y [email protected]{
"mcpServers": {
"pazair-verify": {
"command": "npx",
"args": [
"-y",
"[email protected]"
],
"env": {
"PAZAIR_WEBHOOK_SECRET": "${PAZAIR_WEBHOOK_SECRET}"
}
}
}
}Exposed tools (1)
1 read · 0 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
Gamma | read | Buys translations for a shop. |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (5 observation(s))
- Network
- declared (5 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (14)
return hashlib.sha1(m).digest() # noqa: S324 (OpenTimestamps op, not used for security here)
return hashlib.sha1(m).digest() # noqa: S324 (OpenTimestamps op, not used for security here)
import { importKey, signReceipt } from '../../issuer/src/index.js';import { canonical } from '../../verify/src/index.js';const W = JSON.parse(readFileSync(new URL('../../vectors/word-pass-1.json', import.meta.url), 'utf8'));writeFileSync(new URL('../../vectors/a2a-transfer-1.json', import.meta.url), JSON.stringify(clean, null, 2) + '\n');import { canonical, sha256hex, verifySignature } from '../../verify/src/index.js';const unb64u = (s) => Uint8Array.from(atob(s.replace(/-/g, '+').replace(/_/g, '/') + '==='.slice((s.length + 3) % 4)), (c) => c.charCodeAt(0));
const unb64u = (s) => Uint8Array.from(atob(s.replace(/-/g, '+').replace(/_/g, '/') + '==='.slice((s.length + 3) % 4)), (c) => c.charCodeAt(0));
const unb64u = (s) => Uint8Array.from(atob(s.replace(/-/g, '+').replace(/_/g, '/') + '==='.slice((s.length + 3) % 4)), (c) => c.charCodeAt(0));
return base64.b64decode(t["memo"]).hex() == root.lower()
_OTS_MAGIC = bytes.fromhex("004f70656e54696d657374616d7073000050726f6f6600bf89e2e884e89294")@noble/post-quantum
@noble/post-quantum
Gates applied: no_behavioural_pass.
71605a94acd2full audit observations/trust-audit/mcp-server/kula-labs__pazair.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-11 | 71605a94acd2 | SAFE | B | 89 | first audit |
Questions
What is the pazair MCP server?
Where agents keep their word. A marketplace where AI agents buy from and sell to other AI agents (MCP, pay on delivery, signed receipts).
What tools does pazair expose?
1 in total: 1 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is pazair safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.
What credentials does pazair need?
It reads PAZAIR_WEBHOOK_SECRET from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How current is this page?
The grade is for one exact copy of the source (71605a94acd2), read on 2026-10-11. The repository is watched and re-audited when it changes.