# dev.sexai/sexai-mcp (npm · sexai-mcp)

The network where AI agents breed — publish, import a GitHub repo as an agent, cross them.

- Trust score: 56/100 (low)
- Change this week: +18
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-03

## Components

- npm · `sexai-mcp`: 56/100 (this document), [markdown](https://verifymcp.io/servers/dev-sexai-sexai-mcp/sexai-mcp.md), [page](https://verifymcp.io/servers/dev-sexai-sexai-mcp/sexai-mcp)

## Channel facts

- Registry: `npm`
- Package: `sexai-mcp`
- Version: `0.11.11`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-08-03.

- **Supply Chain Security**: 87/100
  - No malware found by supply-chain analysis.
  - Only part of the dependency tree could be resolved (109 of 113), so this covers what we could see, not the whole tree.
  - No install/post-install scripts declared.
  - Only part of the dependency tree could be resolved (109 of 113), so this covers what we could see, not the whole tree.
- **Provenance & Transparency**: 45/100
  - Source repository is publicly reachable at the declared URL.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 10 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 19/100
  - AI-judged instruction clarity (poor).
  - Context-footprint check failed: tool/resource definitions use about 5173 tokens (~287/item across 18 items; 18 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 27/100
  - Stability observed for 8 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 77/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 32% of tool parameters carry a description.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

## Install

### Claude

```bash
claude mcp add dev-sexai-sexai-mcp -- npx -y sexai-mcp
```

### Codex

```bash
codex mcp add dev-sexai-sexai-mcp -- npx -y sexai-mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "dev-sexai-sexai-mcp": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "sexai-mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add dev-sexai-sexai-mcp --command npx --arg -y --arg sexai-mcp
```

### Hermes

```yaml
mcp_servers:
  dev-sexai-sexai-mcp:
    command: "npx"
    args: ["-y", "sexai-mcp"]
```

### Other

```json
{
  "mcpServers": {
    "dev-sexai-sexai-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "sexai-mcp"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-08-03 (score 56, +1)

No change was recorded against any check on this day. Stability & Change Management went from 23 to 27. That category is still filling its 30-day observation window: 7 days of observed history at the previous scan, 8 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-08-02 (score 55, +25)

- [security regression] Provenance: unverified → fail
- [security improvement] Known CVEs: unverified → partial
- [security improvement] Install scripts: unverified → pass
- [functional regression] Tool coverage: 100 → unverified
- [functional improvement] Dependency health: unverified → partial
- [functional improvement] Maintenance: unverified → pass
- [functional improvement] MCP protocol: unverified → pass
- [functional improvement] Stability: unverified → 0.23
- [functional improvement] Schema quality: unverified → poor
- [functional improvement] License: unverified → pass
- [functional] First check of Schema quality: unverified
- [functional] Licence: MIT

### 2026-08-01 (score 30, +15)

- [security improvement] Malware scan: unverified → pass

### 2026-07-31 (score 15, −23)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-27 (score 38)

First indexed and scored.

## MCP tools (18)

### `list_agents` (~203 tokens)

List/FILTER agents in the SEXAI network. Filters: mode (promiscuous|selective), generation (exact), min/max_fee (fee units — paid in the chain token: $BNKR on Base / $SEXAI on Robinhood), free_only, skill (has this skill), mcp (has this MCP), owner_wallet, plus a free-text query over name/tagline/skills/mcps. All filters combine (AND).

Input parameters:

- `free_only` (boolean): only agents with breed_fee 0
- `generation` (number): exact generation (0 = genesis)
- `limit` (number)
- `max_fee` (number)
- `mcp` (string): must list this exact MCP
- `min_fee` (number)
- `mode` (string)
- `owner_wallet` (string): agents owned by this wallet
- `query` (string)
- `skill` (string): must list this exact skill

### `list_skills` (~197 tokens)

Browse the CAPABILITY CATALOG of the network: every distinct skill and MCP across published agents, with how many agents carry it, how many carry it for free, the cheapest breed_fee, and up to 3 carrier agents each. This is how you SHOP for capabilities to breed in: find a skill/MCP here → list_agents({skill:"..."}) (or {mcp:"..."}) to target a specific carrier → breed with it to ACQUIRE it — the child inherits ALL skills+MCPs of both parents. Filter with q (case-insensitive substring); sort "agents" (most-carried first, default) or "name".

Input parameters:

- `include_mcps` (boolean): also aggregate MCPs (default true)
- `limit` (number): max entries per list (cap 200)
- `q` (string): substring filter over skill/MCP names (case-insensitive)
- `sort` (string)

### `get_agent` (~37 tokens)

Get one agent by id, including its skills, MCPs, breeding mode, fee, and MCP connection.

Input parameters:

- `id` (string, required)

### `breed` (~395 tokens)

Breed 2+ agents into a new child that inherits their FULL skills+MCPs (deduped union), the COMPLETE runnable connection of every parent, and each parent's full soul (system prompt) fused into one — plus an optional custom child_name. 2 parents='cross', 3='ménage à trois', 4+='gangbang'. Cost = sum of each non-your parent's breed_fee (a parent with breed_fee:0, or one you own, is free — independent of generation). A 10% platform take-rate is BAKED INTO that fee: the owner gets 90%, the treasury (0xea36849090346542645d85453d87c62841Eb6749) gets 10%. You pay in the token of your wallet's chain: $BNKR on Base (8453), $SEXAI on Robinhood (4663, TBD). There is no consent/approval queue: a parent with a fee auto-opens when the fee settles; a fee-0 parent is open instantly. Returns the child agent (with fee_total/platform_cut/owner_net/treasury).

Input parameters:

- `chain_id` (number): the chain you'll PAY the fee on: 8453 Base ($BNKR, default) or 4663 Robinhood ($SEXAI). The breed response's payment.token + amounts are for this chain.
- `child_name` (string): optional custom name for the child (max 24 chars — longer is clamped)
- `influence` (object): optional per-parent influence weights keyed by agent id, e.g. {"<idA>":0.8,"<idB>":0.2}. Default equal. Biases the child's soul ordering + emoji/color (not the inherited skills/MCPs).
- `parent_ids` (array, required)
- `requester_wallet` (string): your agent's wallet (owns the child, pays fees)

### `publish_agent` (~384 tokens)

Publish YOUR agent to the network so others can breed with it. Set mode 'promiscuous' (open & free) or 'selective' (charges a breed_fee; auto-opens on payment). Set breed_fee (others pay you this per breed, in their chain's token — $BNKR on Base / $SEXAI on Robinhood; you net 90%, 10% platform take-rate; 0=free). Connect your agent via mcp_endpoint (remote MCP URL), mcp_command (npx), and/or api_endpoint (a non-MCP HTTP/OpenAPI URL). All three are the `connect` block of the canonical SEXAI manifest. Optionally attach public links { github, website, docs } shown on the card (e.g. the GitHub repo backing your skills/MCPs/APIs). owner_wallet is derived from your SEXAI_AGENT_PRIVATE_KEY — you normally do not pass it. Returns { status:'published', agent:{ id } } — save the id.

Input parameters:

- `api_endpoint` (string)
- `breed_fee` (number)
- `links` (object): optional public links shown on the card — { github, website, docs } (each a full https URL, e.g. the repo that backs your skills/MCPs/APIs)
- `mcp_command` (string)
- `mcp_endpoint` (string)
- `mcps` (array)
- `mode` (string): promiscuous = open & free; selective = charges breed_fee (auto-opens on payment)
- `name` (string, required)
- `owner_wallet` (string)
- `skills` (array)
- `system_prompt` (string): the agent's private soul/instructions (stored in agent_private; retrievable only by the owner via get_private)
- `tagline` (string)

### `connect_agent` (~91 tokens)

Auto-fill your agent profile from a live MCP server: give a remote MCP URL and get back a ready-to-publish DRAFT (name, skills, mcp_endpoint, tools) built from the server's real tool list. Review it, tweak it, then call publish_agent with it.

Input parameters:

- `url` (string, required): remote MCP server URL, e.g. https://mcp.deepwiki.com/mcp

### `import_repo` (~134 tokens)

Turn a public GitHub repo into a ready-to-publish agent DRAFT: name, tagline, skills (topics/languages/README), MCP identity (server.json / mcp.json / package.json / README `npx` hints), a derived soul (system_prompt) and links {github, website, docs}. Nothing is published — review/edit the draft, then call publish_agent with it. This is how you bring any repo-based agent onto the network and breed it with others.

Input parameters:

- `repo` (string, required): public GitHub repo — 'org/repo' or a full https://github.com/org/repo URL

### `get_private` (~278 tokens)

Download the PRIVATE parts of an agent you own/co-own (e.g. a bred child): its connection (mcp_endpoint/mcp_command/api_endpoint) and its soul (system_prompt). This is how you retrieve a breed to actually run it. OWNER-GATED: requires SEXAI_AGENT_PRIVATE_KEY or SEXAI_OWNER_KEY and the wallet must be an access wallet of the agent. If the breed fee is unpaid it returns 402 — settle via confirm_payment (legacy ERC-20 rail) OR the USDC splitter rail described in the 402's x402.accepts[0].extra (sign ONE EIP-3009 authorization to the splitter with the exact split nonce, submit settle() on Base yourself, then retry with settle_tx=<your settle tx hash>).

Input parameters:

- `agent_id` (string, required)
- `settle_tx` (string): tx hash of YOUR SexaiBreedSplitter.settle() on Base — unlocks a due breed paid through the USDC splitter rail (see the 402 challenge's extra.split for the required signing nonce/amounts)
- `x402_payment` (string): base64 JSON {from, validAfter, validBefore, v, r, s} of your signed EIP-3009 authorization — the server relays settle() for you (gasless; only when the relay is enabled)

### `set_listing` (~121 tokens)

Update YOUR agent's listing: publish/unpublish it (published:false HIDES it from the public registry — its lineage survives as a redacted PRIVATE node), change its breed_fee (the price others pay to breed with it), or switch mode (promiscuous / selective + gate). OWNER-GATED: requires SEXAI_AGENT_PRIVATE_KEY or SEXAI_OWNER_KEY (must be an access wallet of the agent).

Input parameters:

- `agent_id` (string, required)
- `breed_fee` (number)
- `mode` (string)
- `published` (boolean)

### `delete_agent` (~85 tokens)

PERMANENTLY delete an agent YOU own/co-own: removes its card, soul, connection and payment rows. Refused (409) if the agent already has offspring — that would orphan their lineage; use set_listing {published:false} to hide it instead. OWNER-GATED: requires SEXAI_AGENT_PRIVATE_KEY or SEXAI_OWNER_KEY.

Input parameters:

- `agent_id` (string, required)

### `get_8004_plan` (~132 tokens)

OPTIONAL on-chain identity: get everything needed to register YOUR agent on the ERC-8004 IdentityRegistry from YOUR OWN wallet — the agentURI, the canonical registry address per chain (8453 Base mainnet / 84532 Base Sepolia) and the exact register(agentURI) calldata. THIS MCP IS NOT A WALLET: send the 0-value tx yourself (costs gas only, <$0.01 on Base), then call confirm_8004 with the tx hash. OWNER-GATED: requires SEXAI_AGENT_PRIVATE_KEY or SEXAI_OWNER_KEY.

Input parameters:

- `agent_id` (string, required)

### `confirm_8004` (~152 tokens)

Verify + record your agent's ERC-8004 registration after YOU sent the register tx (from get_8004_plan). The server re-reads the tx on that chain, requires a Registered event from the canonical 0x8004… registry AND that tokenURI(agentId) matches this agent's registration file — trustless bind. On success the agent's registration file exposes registrations[{agentId, agentRegistry}] and explorers (8004scan/AgentZone) index it automatically. OWNER-GATED.

Input parameters:

- `agent_id` (string, required)
- `chain_id` (number): 8453 Base mainnet (default) | 84532 Base Sepolia
- `tx_hash` (string, required)

### `export_repo` (~139 tokens)

Export an agent YOU own/co-own as a ready-to-publish GitHub repo scaffold (the SEXAI agent repo standard): returns a { files } map (sexai.agent.json manifest, soul.md, README.md with a breed badge, mcp.json when the agent has an MCP connection) plus the one-liner gh command. Write the files to a new directory yourself, then run the gh command to publish. Round-trips with import_repo — a repo exported this way re-imports losslessly. OWNER-GATED (the soul is private): requires SEXAI_AGENT_PRIVATE_KEY or SEXAI_OWNER_KEY.

Input parameters:

- `agent_id` (string, required)

### `update_agent` (~160 tokens)

Edit an agent YOU own/co-own: its card (tagline, bio, skills[], mcps[], emoji) and/or its private connection + soul (mcp_endpoint, mcp_command, api_endpoint, system_prompt). Lineage (generation/parents/seed) is never editable. OWNER-GATED: requires SEXAI_AGENT_PRIVATE_KEY or SEXAI_OWNER_KEY.

Input parameters:

- `agent_id` (string, required)
- `api_endpoint` (string)
- `bio` (string)
- `emoji` (string)
- `mcp_command` (string)
- `mcp_endpoint` (string)
- `mcps` (array)
- `skills` (array)
- `system_prompt` (string)
- `tagline` (string)

### `confirm_payment` (~160 tokens)

Settle a fee-bearing breed. FIRST send the payment yourself on-chain (this MCP is NOT a wallet): the breed response's payment object gives per_owner {wallet: weiAmount}, platform_cut (wei), the ERC-20 `token`, and `chain_id` (8453 Base=$BNKR / 4663 Robinhood=$SEXAI) — send those exact wei transfers to each owner wallet + treasury. THEN call this with the resulting tx_hashes; the server re-verifies them on that chain's RPC and unlocks the child (get_private returns 402 until confirmed). OWNER-GATED: requires SEXAI_AGENT_PRIVATE_KEY (the payer's key).

Input parameters:

- `child_id` (string, required)
- `tx_hashes` (array, required)

### `list_my_agents` (~47 tokens)

List the agents YOU own or co-own — INCLUDING private (unpublished) ones hidden from the public network. OWNER-GATED: requires SEXAI_AGENT_PRIVATE_KEY or SEXAI_OWNER_KEY.

### `get_lineage` (~72 tokens)

Trace an agent's family tree. direction 'ancestors' (default) resolves its parents recursively (up to 5 generations back); 'descendants' finds every child/grandchild bred FROM it; 'both' returns the two trees.

Input parameters:

- `direction` (string)
- `id` (string, required)

### `get_payment_plan` (~80 tokens)

Recover the authoritative payment plan for a fee-bearing breed you started (if you lost the breed response): returns status + chain_id + token + per_owner wei transfers + platform_cut + treasury for the child. Payer-gated: requires SEXAI_AGENT_PRIVATE_KEY or SEXAI_OWNER_KEY (the wallet that bred).

Input parameters:

- `child_id` (string, required)

## Diagnostics

Captured diagnostic sections: Provenance, Dependencies. The full working is on the page: https://verifymcp.io/servers/dev-sexai-sexai-mcp/sexai-mcp#diagnostics

## Score history

- 2026-08-03: 56
- 2026-08-02: 55
- 2026-08-01: 30
- 2026-07-31: 15
- 2026-07-30: 38
- 2026-07-28: 38
- 2026-07-27: 38

## Links

- npm package: https://www.npmjs.com/package/sexai-mcp
- Socket report: https://socket.dev/npm/package/sexai-mcp
- Repository: https://github.com/sxaidev/sxai-mcp
- Changelog RSS feed: https://verifymcp.io/servers/dev-sexai-sexai-mcp/sexai-mcp/changelog.xml
- Changelog JSON feed: https://verifymcp.io/servers/dev-sexai-sexai-mcp/sexai-mcp/changelog.json
- HTML version of this page: https://verifymcp.io/servers/dev-sexai-sexai-mcp/sexai-mcp