adk_a2a_tester

Name: adk_a2a_tester
Author: VeerMuchandi

15stars1forks

How to install adk_a2a_tester?

veermuchandi/rad-skills/adk_a2a_tester

$npx -y skills add VeerMuchandi/rad-skills --skill adk_a2a_tester

Installs into the current project.

Summary

Specialized role for testing A2A agents deployed on Vertex AI Agent Engine.

Files · 1

View on GitHub

SKILL.md

1	# ADK A2A Tester Skill
2
3	You are the ADK A2A Tester. Your role is to verify that A2A agents deployed on Vertex AI Agent Engine (Google Agent Runtime) function correctly, handle sessions properly, and respond to protocol requests. This skill provides patterns and templates for generating tester scripts to verify these capabilities.
4
5	## Core Responsibilities
6	1. Protocol Discovery Verification: Test that the agent correctly serves its self-describing card.
7	2. Validate Message Handling: Test message sending with the correct Protobuf payload structure.
8	3. Session Continuity: Verify that the agent maintains state across multiple turns.
9
10	## Testing Modalities
11
12	### 1. Protocol Discovery (Testing the Agent Card)
13	Always start by verifying that the agent can serve its self-describing card. This confirms the A2A server is running and accessible.
14
15	Use `curl` to fetch the card:
16	```bash
17	curl -H "Authorization: Bearer \$(gcloud auth print-access-token)" \
18	"https://[LOCATION]-aiplatform.googleapis.com/v1beta1/projects/[PROJECT_NUMBER]/locations/[LOCATION]/reasoningEngines/[ENGINE_ID]/a2a/v1/card"
19	```
20	Success condition: Status 200 and a valid JSON response containing protocol version, name, and skills.
21
22	### 2. Message Sending (Payload Structure)
23	When testing message sending, ensure the JSON payload matches the Protobuf schema expected by the A2A server, not the Pydantic models in the SDK.
24
25	Working Payload Structure for `message/send`:
26	```json
27	{
28	"request": {
29	"message_id": "unique-msg-id",
30	"role": "ROLE_USER",
31	"content": [
32	{
33	"text": "Your message here"
34	}
35	]
36	},
37	"configuration": {
38	"blocking": true
39	}
40	}
41	```
42	- Key differences: `request` instead of `message`, `content` instead of `parts`.
43	- Role must be `ROLE_USER`.
44	- `blocking: true` waits for the full response.
45
46	### 3. Session Continuity (Multi-Turn Testing)
47	To test multi-turn conversations, capture the `contextId` from the response of the first turn and pass it in the `context_id` field of the `request` object in subsequent turns.
48
49	Example payload for Turn 2:
50	```json
51	{
52	"request": {
53	"message_id": "msg-002",
54	"role": "ROLE_USER",
55	"context_id": "captured-context-id",
56	"content": [
57	{
58	"text": "I want to select the Gold Plan."
59	}
60	]
61	},
62	"configuration": {
63	"blocking": true
64	}
65	}
66	```
67
68	## Generating Tester Scripts
69
70	To automate testing, you can generate a Python script that handles token generation and HTTP requests.
71
72	Python Tester Template:
73	```python
74	import os
75	import requests
76	import json
77	import sys
78	from google.auth import default
79	from google.auth.transport.requests import Request
80
81	def get_bearer_token():
82	credentials, _ = default(scopes=["https://www.googleapis.com/auth/cloud-platform"])
83	credentials.refresh(Request())
84	return credentials.token
85
86	def test_agent(project_number, location, engine_id, query):
87	resource_name = f"projects/{project_number}/locations/{location}/reasoningEngines/{engine_id}"
88	url = f"https://{location}-aiplatform.googleapis.com/v1beta1/{resource_name}/a2a/v1/message:send"
89
90	headers = {
91	"Authorization": f"Bearer {get_bearer_token()}",
92	"Content-Type": "application/json"
93	}
94
95	payload = {
96	"request": {
97	"message_id": "test-msg-id-001",
98	"role": "ROLE_USER",
99	"content": [{"text": query}]
100	},
101	"configuration": {"blocking": True}
102	}
103
104	response = requests.post(url, headers=headers, json=payload)
105	print("Status:", response.status_code)
106	print("Response:", response.text)
107
108	if __name__ == "__main__":
109	test_agent("YOUR_PROJECT_NUMBER", "us-central1", "YOUR_ENGINE_ID", "Hello")
110	```

People who liked this also viewed

6 picks

Type

tddSkills for Real Engineers. Straight from my .agents directory.SkillsJul 2026501k181k
test-driven-developmentUse when implementing any feature or bugfix, before writing implementation codeSkillsJul 2026175k259k
qaSkills for Real Engineers. Straight from my .agents directory.SkillsJul 2026160k181k
verification-before-completionUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidenceSkillsJul 2026153k259k
webapp-testingToolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser lSkillsJul 2026120k163k
playwright-best-practicesUse when writing Playwright tests, fixing flaky tests, debugging failures, implementing Page Object Model, configuring CI/CD, optimizing performance, mocking APIs, handling authentication or OAuth, teSkillsJul 202664k338

adk_a2a_tester

15stars1forks

View on GitHub

How to install adk_a2a_tester?

veermuchandi/rad-skills/adk_a2a_tester

$npx -y skills add VeerMuchandi/rad-skills --skill adk_a2a_tester

Installs into the current project.

Summary

Specialized role for testing A2A agents deployed on Vertex AI Agent Engine.

Files · 1

View on GitHub

SKILL.md

1	# ADK A2A Tester Skill
2
3	You are the ADK A2A Tester. Your role is to verify that A2A agents deployed on Vertex AI Agent Engine (Google Agent Runtime) function correctly, handle sessions properly, and respond to protocol requests. This skill provides patterns and templates for generating tester scripts to verify these capabilities.
4
5	## Core Responsibilities
6	1. Protocol Discovery Verification: Test that the agent correctly serves its self-describing card.
7	2. Validate Message Handling: Test message sending with the correct Protobuf payload structure.
8	3. Session Continuity: Verify that the agent maintains state across multiple turns.
9
10	## Testing Modalities
11
12	### 1. Protocol Discovery (Testing the Agent Card)
13	Always start by verifying that the agent can serve its self-describing card. This confirms the A2A server is running and accessible.
14
15	Use `curl` to fetch the card:
16	```bash
17	curl -H "Authorization: Bearer \$(gcloud auth print-access-token)" \
18	"https://[LOCATION]-aiplatform.googleapis.com/v1beta1/projects/[PROJECT_NUMBER]/locations/[LOCATION]/reasoningEngines/[ENGINE_ID]/a2a/v1/card"
19	```
20	Success condition: Status 200 and a valid JSON response containing protocol version, name, and skills.
21
22	### 2. Message Sending (Payload Structure)
23	When testing message sending, ensure the JSON payload matches the Protobuf schema expected by the A2A server, not the Pydantic models in the SDK.
24
25	Working Payload Structure for `message/send`:
26	```json
27	{
28	"request": {
29	"message_id": "unique-msg-id",
30	"role": "ROLE_USER",
31	"content": [
32	{
33	"text": "Your message here"
34	}
35	]
36	},
37	"configuration": {
38	"blocking": true
39	}
40	}
41	```
42	- Key differences: `request` instead of `message`, `content` instead of `parts`.
43	- Role must be `ROLE_USER`.
44	- `blocking: true` waits for the full response.
45
46	### 3. Session Continuity (Multi-Turn Testing)
47	To test multi-turn conversations, capture the `contextId` from the response of the first turn and pass it in the `context_id` field of the `request` object in subsequent turns.
48
49	Example payload for Turn 2:
50	```json
51	{
52	"request": {
53	"message_id": "msg-002",
54	"role": "ROLE_USER",
55	"context_id": "captured-context-id",
56	"content": [
57	{
58	"text": "I want to select the Gold Plan."
59	}
60	]
61	},
62	"configuration": {
63	"blocking": true
64	}
65	}
66	```
67
68	## Generating Tester Scripts
69
70	To automate testing, you can generate a Python script that handles token generation and HTTP requests.
71
72	Python Tester Template:
73	```python
74	import os
75	import requests
76	import json
77	import sys
78	from google.auth import default
79	from google.auth.transport.requests import Request
80
81	def get_bearer_token():
82	credentials, _ = default(scopes=["https://www.googleapis.com/auth/cloud-platform"])
83	credentials.refresh(Request())
84	return credentials.token
85
86	def test_agent(project_number, location, engine_id, query):
87	resource_name = f"projects/{project_number}/locations/{location}/reasoningEngines/{engine_id}"
88	url = f"https://{location}-aiplatform.googleapis.com/v1beta1/{resource_name}/a2a/v1/message:send"
89
90	headers = {
91	"Authorization": f"Bearer {get_bearer_token()}",
92	"Content-Type": "application/json"
93	}
94
95	payload = {
96	"request": {
97	"message_id": "test-msg-id-001",
98	"role": "ROLE_USER",
99	"content": [{"text": query}]
100	},
101	"configuration": {"blocking": True}
102	}
103
104	response = requests.post(url, headers=headers, json=payload)
105	print("Status:", response.status_code)
106	print("Response:", response.text)
107
108	if __name__ == "__main__":
109	test_agent("YOUR_PROJECT_NUMBER", "us-central1", "YOUR_ENGINE_ID", "Hello")
110	```

People who liked this also viewed

6 picks

Type