rpi-kit 1.0.0

package/commands/rpi/test.md

@@ -0,0 +1,194 @@
+---
+name: rpi:test
+description: Run TDD (Red-Green-Refactor) on a specific task or set of tasks from the plan. Writes one failing test, verifies failure, implements minimal code, verifies pass, refactors.
+argument-hint: "<feature-slug> [--task <id>] [--all]"
+allowed-tools:
+  - Read
+  - Write
+  - Edit
+  - Bash
+  - Glob
+  - Grep
+  - Agent
+  - AskUserQuestion
+---
+
+<objective>
+Execute strict TDD cycles (RED → GREEN → REFACTOR) for tasks in a feature's PLAN.md. One test at a time, verify fail, implement, verify pass.
+</objective>
+
+<process>
+
+## 1. Load config and parse arguments
+
+Read `.rpi.yaml` for configuration (folder, test_runner).
+Parse `$ARGUMENTS`:
+- First argument: `{feature-slug}` (required)
+- `--task <id>`: run TDD for a specific task (e.g., `1.2`)
+- `--all`: run TDD for all uncompleted tasks sequentially
+
+If no `--task` or `--all`, ask the user which task to work on.
+
+## 2. Validate prerequisites
+
+Read `{folder}/{feature-slug}/plan/PLAN.md`. If missing:
+```
+Plan not found. Run /rpi:plan {feature-slug} first.
+```
+
+Read `{folder}/{feature-slug}/plan/eng.md` for technical context (especially Testing Strategy section).
+
+## 3. Detect test infrastructure
+
+Scan the project for existing test setup:
+- Look for test config files: `jest.config.*`, `vitest.config.*`, `pytest.ini`, `pyproject.toml [tool.pytest]`, `*.test.*`, `*.spec.*`
+- Identify the test runner command (from config `test_runner` or auto-detect)
+- Identify test file naming convention (`*.test.ts`, `*.spec.ts`, `*_test.py`, etc.)
+- Identify assertion style (`expect`, `assert`, etc.)
+
+If no test infrastructure found, ask the user:
+"No test setup detected. What test framework and runner should I use?"
+
+## 4. For each target task, run TDD cycle
+
+### Phase RED: Write failing test
+
+Launch test-engineer agent:
+```
+You are the test-engineer agent for the RPI workflow.
+
+Read these files:
+- {folder}/{feature-slug}/plan/PLAN.md
+- {folder}/{feature-slug}/plan/eng.md
+
+Current task:
+**{task_id}** {task_description}
+Files: {files}
+
+Test infrastructure:
+- Framework: {detected_framework}
+- File convention: {convention}
+- Assertion style: {style}
+
+Write ONE failing test for this task:
+1. Create or edit the appropriate test file following project conventions
+2. Write a single test that describes the expected behavior
+3. The test must exercise real code through the public interface
+4. Use clear, behavior-describing test name
+5. Minimal assertions — one logical check
+
+Do NOT write any implementation code. Only the test.
+
+Follow test-engineer rules from RPI agent guidelines.
+```
+
+### Phase VERIFY RED: Confirm correct failure
+
+Run the test:
+```bash
+{test_runner} {test_file}
+```
+
+Check the output:
+- **Test fails with expected reason** (function/module not found, assertion fails) → proceed to GREEN
+- **Test errors** (syntax error, import error, typo) → fix the test, re-run
+- **Test passes** → the behavior already exists. Either the test is wrong or the task is already done. Ask the user.
+
+Report to user:
+```
+RED: Test fails correctly.
+  Test: {test_name}
+  Failure: {failure_reason}
+```
+
+### Phase GREEN: Minimal implementation
+
+Launch plan-executor agent:
+```
+You are implementing a single task using TDD.
+
+The following test is currently FAILING:
+{test_file}:{test_name}
+Failure reason: {failure_reason}
+
+Write the MINIMAL code to make this test pass.
+- Only touch files listed for this task
+- Do NOT add features beyond what the test requires
+- Do NOT write additional tests
+- Match existing code style
+
+Task context:
+**{task_id}** {task_description}
+Files: {files}
+```
+
+### Phase VERIFY GREEN: Confirm pass
+
+Run the test again:
+```bash
+{test_runner} {test_file}
+```
+
+Also run the full test suite to check for regressions:
+```bash
+{test_runner}
+```
+
+Check:
+- **Target test passes** → proceed to REFACTOR
+- **Target test fails** → fix implementation, re-run (do NOT change the test)
+- **Other tests break** → fix regressions before proceeding
+
+Report to user:
+```
+GREEN: Test passes.
+  Test: {test_name}
+  All tests: {pass_count}/{total_count} passing
+```
+
+### Phase REFACTOR: Clean up
+
+Review the implementation just written:
+- Remove duplication
+- Improve names
+- Extract helpers if warranted (3+ uses)
+- Do NOT add new behavior
+
+After refactoring, re-run tests to confirm still green:
+```bash
+{test_runner}
+```
+
+### Commit
+
+If all tests pass:
+```bash
+git add {changed_files}
+git commit -m "{type}({task_id}): {task_description}"
+```
+
+## 5. Check for additional tests needed
+
+After the first TDD cycle for a task, check eng.md for additional test scenarios:
+- Edge cases mentioned in the plan
+- Error handling paths
+- Boundary conditions
+
+If more tests are needed, ask the user:
+"Task {task_id} has additional test scenarios. Continue with next test cycle?"
+
+If yes, repeat the TDD cycle (RED → GREEN → REFACTOR) for each additional test.
+
+## 6. Report results
+
+```
+TDD complete for task {task_id}:
+- Tests written: {N}
+- All passing: {pass_count}/{total_count}
+- Files changed: {list}
+
+{If --all: proceed to next task}
+{If single task: suggest next task or /rpi:implement to continue}
+```
+
+</process>

package/package.json

@@ -0,0 +1,35 @@
+{
+  "name": "rpi-kit",
+  "version": "1.0.0",
+  "description": "Research → Plan → Implement. A systematic feature development workflow for Claude Code.",
+  "license": "MIT",
+  "author": "Daniel Mendes",
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/dmend3z/rpi-kit.git"
+  },
+  "keywords": [
+    "claude-code",
+    "claude",
+    "ai",
+    "workflow",
+    "feature-development",
+    "tdd",
+    "codex",
+    "agents"
+  ],
+  "bin": {
+    "rpi-kit": "./bin/cli.js"
+  },
+  "files": [
+    "bin/",
+    "commands/",
+    "agents/",
+    "AGENTS.md",
+    "README.md",
+    "LICENSE"
+  ],
+  "scripts": {
+    "postinstall": "node bin/cli.js install --silent"
+  }
+}