HAL9000
e4224eb8ec
- Fix malformed JSON parse error: wrap request.json() in try/except, return -32700 Parse error instead of unhandled HTTP 500 - Fix JSON-RPC error responses: add required 'id' field to all error responses per JSON-RPC 2.0 Section 5 - Fix HTTP status codes: return HTTP 200 for all JSON-RPC responses (error codes expressed in JSON body per JSON-RPC 2.0 over HTTP) - Fix error message leakage: return generic messages instead of raw Pydantic ValidationError internals (CWE-209) - Add catch-all exception handler returning -32603 Internal error - Fix Agent Card url field: use base server URL without /a2a suffix; interfaces[0].url correctly uses endpoint URL with /a2a suffix - Fix 0.0.0.0 host: substitute 127.0.0.1 for Agent Card URL - Fix context typing: replace context: Any with context: Context in all step files per project convention - Fix inline imports: move all imports to top of step files - Fix _COMMANDS typing: use dict[str, Callable[[], None]] to avoid type: ignore suppression in helper files - Add BDD scenarios for entity-sync and namespace-mgmt skill enumeration - Update server_lifecycle.feature: correct HTTP status assertions to 200 ISSUES CLOSED: #867
230 lines
10 KiB
Gherkin
230 lines
10 KiB
Gherkin
@phase2 @a2a @agent_card
|
|
Feature: A2A Agent Card Discovery
|
|
As an A2A-compatible client
|
|
I want to discover the agent's capabilities via an Agent Card
|
|
So that I can determine which operations the CleverAgents server supports
|
|
|
|
# -----------------------------------------------------------------------
|
|
# Agent Card model construction
|
|
# -----------------------------------------------------------------------
|
|
|
|
Scenario: Build an Agent Card with a facade
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card name should be "CleverAgents"
|
|
And the agent card version should be "1.0"
|
|
And the agent card should have a non-empty url
|
|
And the agent card should have at least 1 skill
|
|
And the agent card should have at least 1 extension
|
|
|
|
Scenario: Build an Agent Card without a facade
|
|
When I build an agent card without a facade
|
|
Then the agent card name should be "CleverAgents"
|
|
And the agent card version should be "1.0"
|
|
And the agent card should have 0 skills
|
|
|
|
Scenario: Agent Card includes supported versions
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card supported versions should contain "1.0"
|
|
|
|
Scenario: Agent Card includes supported operations
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card should list extension operations
|
|
|
|
Scenario: Agent Card includes security schemes
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card should have at least 1 security scheme
|
|
|
|
Scenario: Agent Card includes protocol interfaces
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card should have at least 1 interface
|
|
And the first interface protocol should be "jsonrpc"
|
|
|
|
Scenario: Agent Card includes default input and output modes
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card default input modes should contain "text"
|
|
And the agent card default output modes should contain "text"
|
|
|
|
Scenario: Agent Card capabilities reflect current state
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card streaming capability should be false
|
|
And the agent card push notifications capability should be false
|
|
|
|
# -----------------------------------------------------------------------
|
|
# Agent Card validation
|
|
# -----------------------------------------------------------------------
|
|
|
|
Scenario: Build rejects invalid host
|
|
When I try to build an agent card with an empty host
|
|
Then a ValueError should be raised for agent card
|
|
|
|
Scenario: Build rejects invalid port zero
|
|
When I try to build an agent card with port 0
|
|
Then a ValueError should be raised for agent card
|
|
|
|
Scenario: Build rejects port above 65535
|
|
When I try to build an agent card with port 70000
|
|
Then a ValueError should be raised for agent card
|
|
|
|
# -----------------------------------------------------------------------
|
|
# Agent Card model validation
|
|
# -----------------------------------------------------------------------
|
|
|
|
Scenario: AgentCard rejects empty name
|
|
When I try to create an AgentCard with empty name
|
|
Then a validation error should be raised for agent card
|
|
|
|
Scenario: AgentCard rejects empty version
|
|
When I try to create an AgentCard with empty version
|
|
Then a validation error should be raised for agent card
|
|
|
|
Scenario: AgentCardSkill rejects empty id
|
|
When I try to create an AgentCardSkill with empty id
|
|
Then a validation error should be raised for agent card
|
|
|
|
Scenario: AgentCardSkill rejects empty name
|
|
When I try to create an AgentCardSkill with empty skill name
|
|
Then a validation error should be raised for agent card
|
|
|
|
Scenario: AgentCardExtension rejects empty uri
|
|
When I try to create an AgentCardExtension with empty uri
|
|
Then a validation error should be raised for agent card
|
|
|
|
Scenario: AgentCardSecurityScheme rejects empty type
|
|
When I try to create an AgentCardSecurityScheme with empty type
|
|
Then a validation error should be raised for agent card
|
|
|
|
Scenario: AgentCardInterface rejects empty protocol
|
|
When I try to create an AgentCardInterface with empty protocol
|
|
Then a validation error should be raised for agent card
|
|
|
|
# -----------------------------------------------------------------------
|
|
# A2A conformance validation
|
|
# -----------------------------------------------------------------------
|
|
|
|
Scenario: Valid Agent Card passes conformance validation
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
And I validate the agent card conformance
|
|
Then the conformance violations should be empty
|
|
|
|
Scenario: Agent Card with empty url fails conformance
|
|
Given an agent card with empty url
|
|
When I validate the agent card conformance
|
|
Then the conformance violations should include "url is required"
|
|
|
|
Scenario: Agent Card with unsupported version fails conformance
|
|
Given an agent card with version "99.0"
|
|
When I validate the agent card conformance
|
|
Then the conformance violations should include "not in supported versions"
|
|
|
|
Scenario: Agent Card with empty supportedVersions fails conformance
|
|
Given an agent card with empty supported versions
|
|
When I validate the agent card conformance
|
|
Then the conformance violations should include "supportedVersions must not be empty"
|
|
|
|
Scenario: Conformance validation rejects non-AgentCard argument
|
|
When I try to validate conformance of a non-AgentCard object
|
|
Then a TypeError should be raised for conformance validation
|
|
|
|
# -----------------------------------------------------------------------
|
|
# Serialization
|
|
# -----------------------------------------------------------------------
|
|
|
|
Scenario: Agent Card serializes to dict
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
And I serialize the agent card to dict
|
|
Then the serialized dict should have key "name"
|
|
And the serialized dict should have key "version"
|
|
And the serialized dict should have key "capabilities"
|
|
And the serialized dict should have key "skills"
|
|
And the serialized dict should have key "supportedVersions"
|
|
And the serialized dict should have key "supportedOperations"
|
|
|
|
Scenario: agent_card_to_dict rejects non-AgentCard argument
|
|
When I try to serialize a non-AgentCard object
|
|
Then a TypeError should be raised for serialization
|
|
|
|
# -----------------------------------------------------------------------
|
|
# Discovery endpoint (ASGI)
|
|
# -----------------------------------------------------------------------
|
|
|
|
Scenario: Discovery endpoint returns Agent Card with skills
|
|
Given a running ASGI test client with facade
|
|
When I request GET /.well-known/agent.json
|
|
Then the response status code should be 200
|
|
And the response body should contain "CleverAgents"
|
|
And the response body should contain "skills"
|
|
And the response body should contain "supportedVersions"
|
|
And the response body should contain "supportedOperations"
|
|
|
|
Scenario: Discovery endpoint Agent Card has correct structure
|
|
Given a running ASGI test client with facade
|
|
When I request GET /.well-known/agent.json and parse JSON
|
|
Then the agent card JSON should have key "name" with value "CleverAgents"
|
|
And the agent card JSON should have key "version" with value "1.0"
|
|
And the agent card JSON "capabilities" should be a dict
|
|
And the agent card JSON "skills" should be a non-empty list
|
|
And the agent card JSON "supportedVersions" should be a non-empty list
|
|
|
|
Scenario: Discovery endpoint Agent Card lists extension operations
|
|
Given a running ASGI test client with facade
|
|
When I request GET /.well-known/agent.json and parse JSON
|
|
Then the agent card JSON "supportedOperations" should be a non-empty list
|
|
And all supported operations should start with "_cleveragents/"
|
|
|
|
# -----------------------------------------------------------------------
|
|
# Skill enumeration from facade
|
|
# -----------------------------------------------------------------------
|
|
|
|
Scenario: Agent Card enumerates plan lifecycle skill
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card skills should include id "plan-lifecycle"
|
|
|
|
Scenario: Agent Card enumerates registry skill
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card skills should include id "registry-crud"
|
|
|
|
Scenario: Agent Card enumerates context skill
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card skills should include id "context-mgmt"
|
|
|
|
Scenario: Agent Card enumerates health skill
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card skills should include id "health-diagnostics"
|
|
|
|
# -----------------------------------------------------------------------
|
|
# Additional conformance edge cases
|
|
# -----------------------------------------------------------------------
|
|
|
|
Scenario: Agent Card with empty defaultInputModes fails conformance
|
|
Given an agent card with empty default input modes
|
|
When I validate the agent card conformance
|
|
Then the conformance violations should include "defaultInputModes must not be empty"
|
|
|
|
Scenario: Agent Card with empty defaultOutputModes fails conformance
|
|
Given an agent card with empty default output modes
|
|
When I validate the agent card conformance
|
|
Then the conformance violations should include "defaultOutputModes must not be empty"
|
|
|
|
Scenario: Agent Card enumerates entity sync skill
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card skills should include id "entity-sync"
|
|
|
|
Scenario: Agent Card enumerates namespace management skill
|
|
Given an A2aLocalFacade for agent card testing
|
|
When I build an agent card from the facade
|
|
Then the agent card skills should include id "namespace-mgmt"
|