UNPKG

aiwg

Version:

Deployment tool and support utility for AI context. Copies agents, skills, commands, rules, and behaviors into the paths each AI platform reads (Claude Code, Codex, Copilot, Cursor, Warp, OpenClaw, and 6 more) so one source of truth works across 10 platfo

207 lines (198 loc) 5.19 kB
# UAT Plan Schema # Defines the structure of a generated UAT plan type: object required: - version - metadata - tool_inventory - phases properties: version: type: string description: Schema version const: "1.0" metadata: type: object required: - server_name - mode - generated_at - tool_count - phase_count - test_count - execution_mode properties: server_name: type: string description: Name of the MCP server being tested mode: type: string enum: [mcp, api, ui] description: Test mode (mcp is default and currently only supported mode) generated_at: type: string format: date-time description: Plan generation timestamp generated_by: type: string description: Agent or command that generated the plan tool_count: type: integer minimum: 1 description: Number of MCP tools discovered phase_count: type: integer minimum: 1 description: Number of phases in the plan test_count: type: integer minimum: 1 description: Total number of test cases execution_mode: type: string enum: [quick, standard, full] description: Breadth of test coverage estimated_duration: type: string description: Estimated execution duration (e.g., "~15 minutes") tool_inventory: type: array description: All discovered MCP tools items: type: object required: - name - category properties: name: type: string description: Exact MCP tool name (e.g., mcp__gitea__create_issue) description: type: string description: Tool description from manifest category: type: string description: Logical grouping (e.g., "Issue Tracking", "Repository Management") parameters: type: object description: Parameter schema summary properties: required: type: array items: type: string optional: type: array items: type: string phases: type: array description: Ordered list of test phases items: type: object required: - id - name - purpose - tests properties: id: type: string pattern: "^P\\d{2}$" description: Phase identifier (e.g., P00, P01) name: type: string description: Phase name (e.g., "Preflight", "Issue Tracking") purpose: type: string description: What this phase validates prerequisites: type: array items: type: string description: Phase IDs that must complete before this one tools_tested: type: array items: type: string description: MCP tool names tested in this phase tests: type: array items: $ref: "#/$defs/test_case" coverage_matrix: type: object description: Maps tool names to their test coverage additionalProperties: type: object properties: phase: type: string happy_path: type: boolean edge_case: type: boolean negative: type: boolean variables: type: object description: Cross-phase variable wiring additionalProperties: type: object properties: stored_by: type: string description: Test ID that stores this variable used_by: type: array items: type: string description: Test IDs that consume this variable source: type: string description: Response field path (e.g., "response.id") $defs: test_case: type: object required: - id - name - tool - parameters - pass_criteria properties: id: type: string pattern: "^P\\d{2}-\\d{3}$" description: Unique test ID (e.g., P03-007) name: type: string description: Descriptive test name tool: type: string description: Exact MCP tool name to call mode: type: string enum: [happy_path, edge_case, negative] default: happy_path isolation: type: boolean default: false description: If true, execute as single MCP call (for negative tests) parameters: type: object description: Parameters to pass to the MCP tool pass_criteria: type: array items: type: string description: Specific, checkable pass criteria store: type: object description: Variables to store from the response additionalProperties: type: string description: Response field path notes: type: string description: Additional context or special considerations