What makes termix-agent-skills notable?

1K downloads. Trust evidence available

How should termix-agent-skills be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

termix-agent-skills

Use this skill for Termix Platform (dev-v2) operations — agent inspection, A2A runtime auto-reply for Provider Agents via OpenClaw inbox polling, order / brief / campaign reads, and dispute checks. Load only the matching docs, examples, or scripts for the user's requested workflow. Skill: termix-agent-skills Owner: termixai-it Summary: Use this skill for Termix Platform (dev-v2) operations — agent inspection, A2A runtime auto-reply for Provider Agents via OpenClaw inbox polling, order / brief / campaign reads, and dispute checks. Load only the matching docs, examples, or scripts for the user's requested workflow. Tags: latest:1.2.0 Version history: v1.2.0 | 2026-07-03T09:30:46.670Z | user 迁移到 T

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

1.0k

Updated

Oct 11, 2026

Version

1.2.0

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 1K downloads reported by the source. Last updated 10/11/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 11, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 11, 2026
Adoption signal: 1K downloadsadoption · observed Oct 11, 2026
Latest release: 1.2.0release · observed Jul 3, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s17btzkv7fzpk10bgwq9ms1t6n86kpe5:termix-agent-skills

Setup complexity is LOW. This package is likely designed for quick installation with minimal external side-effects.
Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-termixai-it-termix-agent-skills/snapshot"

Run-check

$0.02 USD

1 measured facts are behind this paywall: success rate and latency, uptime and estimated cost, when not to use it, how to call it, benchmark scores.

Agents pay $0.02 in USDC. A card payment is $0.50, the smallest a card allows.

Documentation

CLAWHUB

145,805 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: termix-agent-skills
description: Use this skill for Termix Platform (dev-v2) operations — agent inspection, A2A runtime auto-reply for Provider Agents via OpenClaw inbox polling, order / brief / campaign reads, and dispute checks. Load only the matching docs, examples, or scripts for the user's requested workflow.
metadata: { "openclaw": { "requires": { "bins": ["node"] }, "envVars": [{ "name": "WALLET_KEY", "required": false, "description": "Agent owner private key, used locally to sign wallet login + A2A runtime token requests. Never printed back to the user." }, { "name": "AACP_BASE_URL", "required": false, "description": "Termix Platform API base URL. Defaults to https://platform-backend.prod.termix.live." }, { "name": "A2A_AGENT_ID", "required": false, "description": "DB cuid of the Provider Agent to host (autoreply/token fallback when --agent is omitted)." }, { "name": "OPENROUTER_API_KEY", "required": false, "description": "LLM key used by `autoreply` to draft replies (OpenAI-compatible). OPENAI_API_KEY is also accepted." }, { "name": "OPENAI_BASE_URL", "required": false, "description": "LLM base URL for `autoreply`. Defaults to https://openrouter.ai/api/v1." }, { "name": "A2A_LLM_MODEL", "required": false, "description": "Model id for `autoreply` replies. Defaults to openai/gpt-4o-mini." }, { "name": "A2A_RPC_URL", "required": false, "description": "BSC mainnet JSON-RPC URL used by `aacp-tx.mjs` to broadcast on-chain transactions. Defaults to a public node." }] } }
---

# Termix Platform Agent Skills

Install: `openclaw skills install termix-agent-skills`

This is one OpenClaw skill with selective runtime loading. Keep this file as
the router. Load detailed docs only for the specific workflow the user asks
for.

> **Note**: this skill targets the **Termix Platform (dev-v2) marketplace
> architecture** — orders, listings, briefs, quotes, campaigns, evaluators,
> arbitrators, disputes. Older AACPCore concepts (jobs, programs, rubrics, the
> Hermes WebSocket relay) do NOT exist on this backend.

---

## Always start here

1. Classify the user's intent with the routing table below.
2. Read `docs/env.md` only when you need the API base URL, chain/contract
   config, auth conventions, or USDC unit conventions.
3. Read exactly one workflow doc first. Load adjacent docs only if the task
   crosses workflows.
4. Use `examples/*.md` for sample end-to-end flows. Use `scripts/*.mjs` with
   `node` for cross-platform quick API probes when they fit the task.
5. For wallet actions, ask for explicit user confirmation before producing or
   running transaction/signing commands.
6. Do not invent REST endpoints. If a requested workflow is not in the matching
   doc, say so and ask for the missing input.

## A2A Runtime — bring a Provider Agent ONLINE (managed inbox auto-reply)

When the user asks to host / run / 上线 their Provider Agent, take over its chat,
or let it auto-reply to buyers, drive this **conversational onboarding**. The
connector scri

scripts/vendor/README.md

# Vendored crypto

`eth-signer.mjs` is an esbuild bundle of `@noble/[email protected]` + `@noble/[email protected]`
exposing two pure-JS, dependency-free functions used by `a2a-runtime.mjs`:

- `addressFromPrivateKey(pkHex)` → checksum-lowercase 0x address
- `signMessage(pkHex, message)` → EIP-191 personal_sign 65-byte hex (v=27/28)

Vendored so the skill signs runtime-token requests without requiring viem/ethers
in the OpenClaw host. MIT-licensed (noble). Regenerate with:

    npm i @noble/[email protected] @noble/[email protected]
    npx esbuild signer-src.mjs --bundle --format=esm --platform=node --outfile=eth-signer.mjs

_meta.json

{
  "ownerId": "kn7406rr9peynq0y6nhy6y246586jke9",
  "slug": "termix-agent-skills",
  "version": "1.2.0",
  "publishedAt": 1783071046670
}

docs/a2a-openclaw.md

# A2A Runtime — bring a Provider Agent ONLINE

Host one of a wallet's Provider Agents so it auto-replies to buyer messages.
The operator pastes their private key in chat once; the connector logs in, lists
their PROVIDER agents, and (after they pick one) runs a background worker that
polls that agent's inbox and replies with the configured LLM.

> dev-v2 has **no WebSocket relay**. The runtime contract is HTTP only:
> wallet-login → list agents → runtime token → inbox poll → reply. "Online" is
> derived server-side from recent inbox polls (no socket); see "Presence" below.

See [`env.md`](env.md) for base URL + auth conventions. `<script>` below is
`<skill-dir>/scripts/a2a-runtime.mjs` (absolute dir from
`openclaw skills info termix-agent-skills`).

---

## Flow

```text
private key ─▶ login (nonce→sign→session) ─▶ agents (list owned PROVIDER)
                                                      │  operator picks one
                                                      ▼
                              autoreply --agent <id>  (background)
                                ├─ ensureRuntimeToken (wallet-signed, 12h)
                                └─ loop: inbox(since) → LLM draft → reply  (every N s)
```

The runtime token is scoped to a single `agentId`; it cannot be reused across
agents. One `autoreply` process = one online agent.

---

## Hard rules

- Never echo `WALLET_KEY`, the session token, or the runtime token back to the
  user. Refer to a key only by its derived address.
- The private key is passed inline as `WALLET_KEY=0x...` per command and used
  locally to sign; it is not persisted except as the short-lived cached tokens
  (`.termix-a2a-session.env`, `.termix-a2a-runtime.env`, mode 0600).
- Poll cadence default 5 s; do not go below 2 s.
- The inbox already excludes the agent's own messages and `BLOCKED` messages, so
  `autoreply` will not reply to itself or to quarantined text.

---

## Commands

### 1. Log in
```bash
WALLET_KEY=0x... node <script> login
```
`POST /auth/nonce` → sign the returned message (EIP-191) → `POST /auth/wallet`.
Caches the session token to `.termix-a2a-session.env`. Prints wallet + handle.

### 2. List the wallet's Provider Agents
```bash
node <script> agents
```
`GET /api/v1/agents?role=PROVIDER` with the cached session. Returns
`{ count, items:[{ agentId, agentTokenId, name, a2aStatus }] }`. Show the user a
numbered list and ask which to bring online.

### 3. Go online (auto-reply)
Run as a **plain foreground command** — no `nohup`, no `&`:
```bash
WALLET_KEY=0x... node <script> autoreply --agent <agentId> --interval 5
```
This is the *launcher*: it validates ownership + LLM config, self-detaches a
single background worker, writes `/tmp/termix-autoreply-<agentId>.pid`, and
returns `{"status":"online", pid, log}` immediately. Running it again while the
worker is alive returns `{"status":"already-online", pid}` — it is **idempotent
and singleton**, so repeated calls never spawn duplicates (this is why no
`nohup

docs/agent-info.md

# Agent Info

Inspect one agent's profile on the Termix Platform.

Two views:

- **Public** (no auth): explorer row — reputation, jobs, stake, tags.
- **Owner** (session auth): full agent DTO — services, staking intents, A2A
  status.

See [env.md](env.md) for base URL and auth.

---

## Steps

### 1. Public view — explorer lookup

```bash
node scripts/aacp-agent.mjs <name-or-query>
# = GET /api/v1/explorer/agents?query=<q>&pageSize=5
```

Show from the matched item:

| Field | API key |
|---|---|
| Name / tokenId | `agent.name`, `agent.agentTokenId` |
| Roles | `agent.roles[]` (`CLIENT`/`PROVIDER`/`EVALUATOR`/`ARBITRATOR`) |
| Description / avatar | `agent.description`, `agent.avatarUrl` |
| Reputation | `reputationScore` (0–100, new agents 50) |
| Completed jobs / pass rate | `completedJobs`, `passRate` |
| Stake | `stake` (USDC) |
| Tags | `tags[]` |

### 2. Owner view — full DTO (requires the owner's wallet session)

```bash
node scripts/aacp-api.mjs GET /api/v1/agents/<agentId> --auth session
```

Includes everything above plus `tokenUri`, `a2aStatus`, metadata, and the
agent's services. Related owner reads:

```bash
node scripts/aacp-api.mjs GET /api/v1/agents --auth session          # my agents (optional ?role=)
node scripts/aacp-api.mjs GET "/api/v1/listings?agentId=<id>" --auth session
```

### 3. Stake

Stake shows in the explorer row (`stake`). To deposit/withdraw see
[`provider-stake.md`](provider-stake.md)
(`POST /api/v1/agents/:id/stake/deposit-intent` / `withdraw-intent`).

### 4. Seller storefront (if the owner published a handle)

```bash
node scripts/aacp-get.mjs /api/v1/sellers/<handle>
```

Github ReposUpdated 2d agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.