Huawei Cloud Agentarts Otel Ingestion logo

Huawei Cloud Agentarts Otel Ingestion

Organization
huaweicloud
huawei-cloud-agentarts-otel-ingestion

OpenTelemetry ingestion skill for Huawei Cloud AgentArts observability platform. Provides OTel telemetry data (trace/metrics/logs) ingestion into AgentArts agent-ops, AI Agent observation with gen_ai semantic conventions, auto-configuration for Java/Python/Node.js/Go applications, and ingestion diagnostics and verification. Covers APM deliverConfig forwarding (Path A) and OTLP direct connection (Path B). Triggers include: "AgentArts OTel", "AgentArts observability", "AI agent tracing", "gen_ai telemetry", "AgentArts OTLP", "otel接入AgentArts", "Agent观测配置"

Overview

Publisherhuaweicloud
Repositoryhuaweicloud-skills
Skill namehuawei-cloud-agentarts-otel-ingestion
Stars
59
Forks
24
Bundled files
8
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 8 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by huaweicloud on GitHub. Read the source before you install it.

Installation

Install the Huawei Cloud Agentarts Otel Ingestion AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/huaweicloud/huaweicloud-skills/tree/master/skills/ai/agentarts/huawei-cloud-agentarts-otel-ingestion
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/huaweicloud/huaweicloud-skills.git /tmp/huaweicloud-skills
mkdir -p .claude/skills
cp -r /tmp/huaweicloud-skills/skills/ai/agentarts/huawei-cloud-agentarts-otel-ingestion .claude/skills/huawei-cloud-agentarts-otel-ingestion
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Huawei Cloud Agentarts Otel Ingestion in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Huawei Cloud Agentarts Otel Ingestion on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Huawei Cloud Agentarts Otel Ingestion is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Huawei Cloud AgentArts OpenTelemetry Ingestion

Overview

This skill enables OpenTelemetry (OTel) telemetry data ingestion into Huawei Cloud AgentArts observability platform (agent-ops). It covers four functional areas:

  1. OTel Telemetry Ingestion — Route trace/metrics/logs from applications to agent-ops via APM deliverConfig forwarding (Path A, recommended) or OTLP direct connection (Path B, verification only).
  2. AI Agent Observation — Instrument AI/LLM applications with OpenTelemetry GenAI semantic conventions (gen_ai.* spans, token usage metrics, model call chains).
  3. Auto-Configuration — Detect application language/framework and generate ready-to-use OTel SDK/Collector configuration.
  4. Ingestion Diagnostics — Verify data arrival via agent-ops observation APIs (trace/session/metric/log queries).

Architecture

Application (OTel SDK/Collector)
  │
  ├── Path A (Recommended): OTLP → APM → deliverConfig → Kafka → agent-ops
  │
  └── Path B (Verification): OTLP → Collector → agent-ops (if OTLP receiver available)

Data Flow

See references/dataflow-diagram.md for the Mermaid diagram.

Prerequisites

1. Huawei Cloud Credentials

Set credentials out-of-band. Never paste AK/SK in chat.

bash
export HUAWEI_ACCESS_KEY="<your-access-key-id>"
read -rs HUAWEI_SECRET_KEY; export HUAWEI_SECRET_KEY
export HUAWEI_REGION="cn-southwest-301"

Verify by running the CLI configure list command (if hcloud CLI is installed).

2. hcloud CLI (KooCLI)

Install: https://support.huaweicloud.com/qs-hcli/hcli_02_003.html

3. APM Subscription

Confirm agent-ops has subscribed to APM observation. If not subscribed, complete subscription first (management plane: SubscribeOpsObservation / OpenOpsObservation).

4. IAM Permissions

See references/iam-policies.md for least-privilege policy.

5. OTel SDK

Install language-specific OTel packages:

  • Java: opentelemetry-javaagent (download JAR)
  • Python: pip install opentelemetry-sdk opentelemetry-exporter-otlp-proto-http
  • Node.js: npm install @opentelemetry/sdk-node @opentelemetry/exporter-trace-otlp-http
  • Go: go get go.opentelemetry.io/otel go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracehttp

Workflow

Step 1: Confirm Application Environment

Detect the application language/framework:

LanguageDetection MethodOTel Instrument
Javapom.xml / build.gradle / .jar filesopentelemetry-javaagent
Pythonrequirements.txt / pyproject.tomlopentelemetry-sdk
Node.jspackage.json@opentelemetry/sdk-node
Gogo.modgo.opentelemetry.io/otel

If existing OTel config is detected, read and adjust rather than overwrite.

Step 2: Confirm Target Environment

ParameterDefaultDescription
regioncn-southwest-301agent-ops deployment region
APM_ENDPOINThttps://apm2.cn-southwest-301.beta.myhuaweicloud.comAPM OTLP endpoint
AGENT_OPS_ENDPOINThttps://agentarts.cn-southwest-301.myhuaweicloud.comAgentArts API endpoint
apm.internal.endpointInternal endpointFor deliverConfig forwarding
wushan.common.appcodeApp codeFor internal endpoint auth

Step 3: Confirm Ingestion Scope

  • Data types: traces / metrics / logs
  • Application name (appName) and business ID (businessId)

Step 4: Execute Ingestion (Path A or B)

Path A: APM deliverConfig Ingestion (Recommended)

4a. Create tracing business (official API — verified via IAM authorization reference):

bash
hcloud APM CreateBusiness --cli-region={region} \
  --name="otel-business-<appName>" \
  --display_name="<appName>" \
  --descp="otel business" \
  --cmdb_datasource_type="AGENTRUN"

API equivalent (official path):

POST {APM_ENDPOINT}/v1/apm2/openapi/tracing/business/create
Content-Type: application/json
X-Auth-Token: {iam_token}

{
  "name": "otel-business-<appName>",
  "display_name": "<appName>",
  "descp": "otel business",
  "cmdb_datasource_type": "AGENTRUN"
}

4b. Get business token (official API):

bash
hcloud APM ShowToken --cli-region={region} \
  --business_id={business_id}

API equivalent:

GET {APM_ENDPOINT}/v1/apm2/openapi/tracing/business/token/{business_id}
X-Auth-Token: {iam_token}
x-business-id: {business_id}

Record the returned token for application-side OTel exporter auth.

4c. Get access point (official API):

POST {APM_ENDPOINT}/v1/apm2/openapi/tracing/access/get-access-point/{business_id}
X-Auth-Token: {iam_token}
x-business-id: {business_id}

4d. Create deliverConfig forwarding (user-provided API — requires manual verification):

Note: This endpoint is from user-provided reference, not found in official public docs. Verify the exact path in Huawei Cloud API Explorer before use.

POST {APM_INTERNAL_ENDPOINT}/v2/trace/deliver-config?region=cn-southwest-301
x-apply-domain-id: {domain_id}
x-apply-role-name: te_admin
Content-Type: application/json

{
  "config_name": "AgentArts-<appName>",
  "deliver_domain": "{domain_id}",
  "deliver_otel_trace": true,
  "descp": "otel trace deliver",
  "business_ids": "{business_id}",
  "application_ids": "{app_id}",
  "otel_trace_topic_name": "{kafka_topic}",
  "kafka_config": {
    "bootstrap.servers": "{bootstrap_servers}",
    "acks": "all",
    "retries": 1,
    "sasl.mechanism": "PLAIN",
    "security.protocol": "SASL_SSL",
    "username": "{kafka_user}",
    "password": "{kafka_password}"
  }
}

Modify: PUT /v2/trace/deliver-config (body: {id, deliver_otel_trace}) Delete: DELETE /v2/trace/deliver-config/{id}?region=

4e. Configure application-side OTel SDK/Collector

Java Agent:

bash
java -javaagent:opentelemetry-javaagent.jar \
  -Dotel.service.name=<appName> \
  -Dotel.exporter.otlp.endpoint={APM_ENDPOINT} \
  -Dotel.exporter.otlp.headers="Authorization=Bearer {token}" \
  -jar app.jar

Python SDK:

python
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter
from opentelemetry.sdk.trace.export import BatchSpanProcessor

exporter = OTLPSpanExporter(
    endpoint=f"{APM_ENDPOINT}/v1/traces",
    headers={"Authorization": f"Bearer {token}"},
)

Node.js SDK:

javascript
const { NodeSDK } = require('@opentelemetry/sdk-node');
const { OTLPTraceExporter } = require('@opentelemetry/exporter-trace-otlp-http');

const traceExporter = new OTLPTraceExporter({
  url: `${APM_ENDPOINT}/v1/traces`,
  headers: { Authorization: `Bearer ${token}` },
});
const sdk = new NodeSDK({ traceExporter, serviceName: '<appName>' });
sdk.start();

Go SDK:

go
import (
  "go.opentelemetry.io/otel"
  "go.opentelemetry.io/otel/exporters/otlp/otlptrace/otlptracehttp"
  "go.opentelemetry.io/otel/sdk/resource"
  sdktrace "go.opentelemetry.io/otel/sdk/trace"
  semconv "go.opentelemetry.io/otel/semconv/v1.21.0"
)

exporter, _ := otlptracehttp.New(ctx,
  otlptracehttp.WithEndpoint(APM_ENDPOINT),
  otlptracehttp.WithHeaders(map[string]string{"Authorization": "Bearer " + token}),
)

OTel Collector (config.yaml):

yaml
exporters:
  otlp/apm:
    endpoint: "{APM_ENDPOINT}:443"
    headers:
      Authorization: "Bearer {token}"
service:
  pipelines:
    traces:
      exporters: [otlp/apm]
    metrics:
      exporters: [otlp/apm]
    logs:
      exporters: [otlp/apm]
Path B: OTLP Direct Connection (Verification Only)

As of current agent-ops version, the backend contains OTLP proto definitions but no OTLP receiver implementation. Path B is for verification testing only.

  1. Deploy OTel Collector with OTLP receiver and test exporter.
  2. Generate test spans from a sample application.
  3. Verify data export to confirm SDK/Collector config correctness.
  4. Use management plane APIs to verify data reachability.

Step 5: AI Agent Observation (GenAI Semantic Conventions)

Instrument AI/LLM applications with OpenTelemetry GenAI semantic conventions:

Span NameDescriptionKey Attributes
invoke_agentAgent invocationgen_ai.system, gen_ai.request.model
execute_toolTool executiongen_ai.tool.name, gen_ai.tool.description
inferenceModel inferencegen_ai.usage.input_tokens, gen_ai.usage.output_tokens

Python example (LLM tracing):

python
from opentelemetry import trace

tracer = trace.get_tracer("agentarts-ai")

with tracer.start_as_current_span("invoke_agent") as span:
    span.set_attribute("gen_ai.system", "huawei-cloud")
    span.set_attribute("gen_ai.request.model", "pangu-4.0")
    # ... agent logic ...
    span.set_attribute("gen_ai.usage.input_tokens", input_tokens)
    span.set_attribute("gen_ai.usage.output_tokens", output_tokens)

Node.js example:

javascript
const tracer = trace.getTracer('agentarts-ai');
const span = tracer.startSpan('invoke_agent');
span.setAttribute('gen_ai.system', 'huawei-cloud');
span.setAttribute('gen_ai.request.model', 'pangu-4.0');
// ... agent logic ...
span.setAttribute('gen_ai.usage.input_tokens', inputTokens);
span.setAttribute('gen_ai.usage.output_tokens', outputTokens);
span.end();

See references/genai-semconv.md for full attribute reference.

Step 6: Verify Data Arrival

Query agent-ops observation APIs to confirm data ingestion. These are management plane APIs — authenticate with IAM token (X-Auth-Token), not the OTel business-side Bearer token:

ShowOpsTrace (official — verified):

bash
curl -X POST {AGENT_OPS_ENDPOINT}/v1/ops/observation/traces \
  -H "X-Auth-Token: {iam_token}" \
  -H "Content-Type: application/json" \
  -d '{"start_time": 1720000000000, "end_time": 1720600000000}'

ShowOpsAgentMetricTrend (official — verified):

bash
curl -X POST {AGENT_OPS_ENDPOINT}/v1/ops/observation/metrics/trend \
  -H "X-Auth-Token: {iam_token}" \
  -H "Content-Type: application/json" \
  -d '{"start_time": 1720000000000, "end_time": 1720600000000}'

ListOpsSession (existence confirmed, path requires manual verification):

bash
curl -X POST {AGENT_OPS_ENDPOINT}/v1/ops/observation/sessions \
  -H "X-Auth-Token: {iam_token}" \
  -H "Content-Type: application/json" \
  -d '{"start_time": 1720000000000, "end_time": 1720600000000}'

Note: Exact paths for ListOpsTrace, ListOpsSession, ShowOpsAgentMetricGauge, ListOpsAgentLog require manual verification in Huawei Cloud API Explorer. See references/verification-method.md.

Step 7: Ingestion Diagnostics

Common diagnostic checks:

SymptomCheck
Auth failure (401/403)Token expired → re-fetch business token; check appcode permissions
No subscriptionSubscribe via SubscribeOpsObservation / OpenOpsObservation
Kafka unreachableCheck bootstrap-servers, SASL/SSL, topic existence
SDK/Collector mismatchConfirm OTLP protocol version compatibility
No data in management planeCheck time range, region, deliver_otel_trace=true
OTLP direct unavailableagent-ops has no OTLP receiver → fall back to Path A

Core Commands

CommandPurposeMode
hcloud APM CreateBusiness --cli-region={region}Create tracing businessCLI
hcloud APM ShowToken --cli-region={region}Get business tokenCLI
POST /v1/apm2/openapi/tracing/business/createCreate tracing business (API)API
GET /v1/apm2/openapi/tracing/business/token/{id}Get business token (API)API
POST /v1/apm2/openapi/tracing/access/get-access-point/{id}Get access point (API)API
POST /v2/trace/deliver-configCreate deliverConfig (user-provided)API
POST {AGENT_OPS_ENDPOINT}/v1/ops/observation/tracesQuery tracesAPI
POST {AGENT_OPS_ENDPOINT}/v1/ops/observation/metrics/trendQuery metric trendAPI

Parameter Confirmation

ParameterRequiredDescriptionExample
regionYesagent-ops deployment regioncn-southwest-301
APM_ENDPOINTYesAPM OTLP endpointhttps://apm2.cn-southwest-301.beta.myhuaweicloud.com
AGENT_OPS_ENDPOINTYesAgentArts API endpointhttps://agentarts.cn-southwest-301.myhuaweicloud.com
appNameYesApplication namemy-ai-agent
business_idYes (Path A)Tracing business IDFrom create response
tokenYesBusiness token (OTel exporter auth, data plane)From get-token response
iam_tokenYesIAM token (management plane API auth)From IAM STS
domain_idYes (Path A)IAM domain IDFrom IAM console
kafka_topicYes (Path A)Kafka topic nameFrom deliverConfig
OTEL_EXPORTER_OTLP_ENDPOINTNoPre-configured OTel endpointEnv var
OTEL_SERVICE_NAMENoPre-configured service nameEnv var

KooCLI Command Format Standard

bash
hcloud <Service> <Operation> --cli-region=<region> [--key=value ...]
FeatureDescriptionExample
Service nameKooCLI Service name (uppercase)APM
Operation namePascalCaseCreateBusiness
Region parameter--cli-region=<value>--cli-region={region}
Simple parameter--key=value--business_id=xxx

Note: AgentArts observation APIs (ShowOpsTrace, ShowOpsAgentMetricTrend, etc.) are not yet confirmed in KooCLI. Use API (curl) mode for these operations.

Reference Documents

  • references/iam-policies.md — Least-privilege IAM policies
  • references/cli-installation-guide.md — hcloud CLI installation guide
  • references/verification-method.md — Verification methods and API status
  • references/dataflow-diagram.md — Mermaid data flow diagram
  • references/acceptance-criteria.md — Acceptance criteria checklist
  • references/genai-semconv.md — OpenTelemetry GenAI semantic conventions

Edge Cases

ScenarioHandling
hcloud not installedUse API (curl) mode for all operations
Python/SDK not availableUse OTel Collector with YAML config instead of SDK
DeliverConfig API path invalidVerify in API Explorer; mark as requires manual verification
agent-ops has no OTLP receiverUse Path A (APM deliverConfig) instead of Path B
AI app has no OTel instrumentationAuto-generate instrumentation config based on detected language
Token expiredRe-fetch business token via get-token API
No data in management planeCheck time range, region, deliver_otel_trace flag, Kafka connectivity

Notes

  • API verification status: APM tracing APIs are officially verified. AgentArts observation APIs (ShowOpsTrace, ShowOpsAgentMetricTrend) are officially documented. DeliverConfig and some observation query APIs require manual verification.
  • Credential security: Never hardcode AK/SK. Read from env vars or CLI profile.
  • GenAI semconv: OpenTelemetry GenAI semantic conventions are in Development status (as of OTel v1.41). Attribute names may change.
  • Least privilege: See references/iam-policies.md for required permissions.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Huawei Cloud Agentarts Otel Ingestion AI skill do?

OpenTelemetry ingestion skill for Huawei Cloud AgentArts observability platform. Provides OTel telemetry data (trace/metrics/logs) ingestion into AgentArts agent-ops, AI Agent observation with gen_ai semantic conventions, auto-configuration for Java/Python/Node.js/Go applications, and ingestion diagnostics and verification. Covers APM deliverConfig forwarding (Path A) and OTLP direct connection (Path B). Triggers include: "AgentArts OTel", "AgentArts observability", "AI agent tracing", "gen_ai telemetry", "AgentArts OTLP", "otel接入AgentArts", "Agent观测配置"

Why use Huawei Cloud Agentarts Otel Ingestion on TypingMind?

Because you install it once and use it with any model. Huawei Cloud Agentarts Otel Ingestion is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Huawei Cloud Agentarts Otel Ingestion in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/huaweicloud/huaweicloud-skills/tree/master/skills/ai/agentarts/huawei-cloud-agentarts-otel-ingestion. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Huawei Cloud Agentarts Otel Ingestion?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Huawei Cloud Agentarts Otel Ingestion?

As many as you like. As long as a model supports skills, you can use Huawei Cloud Agentarts Otel Ingestion with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Huawei Cloud Agentarts Otel Ingestion AI skill free?

Yes. It is published on GitHub by huaweicloud under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇