slack-auth-1770277653/.claude/commands/implement-feature.md
jordan 0a1a1e4b3d
Some checks failed
ci/woodpecker/push/woodpecker Pipeline failed
ci/woodpecker/manual/woodpecker Pipeline was successful
Initialize project from skeleton template
2026-02-05 07:47:34 +00:00

101 lines
2.2 KiB
Markdown

---
description: Implement all tasks for a feature autonomously
argument-hint: <feature-slug>
allowed-tools: Bash, Read, Write, Edit, Glob, Grep, Task
---
Implement all tasks for feature: $ARGUMENTS
## Instructions
### 1. Load Feature Context
```bash
sdlc feature show $ARGUMENTS --json
```
Parse the output to understand current phase, existing artifacts, and task status.
### 2. Verify Prerequisites
Check that these artifacts exist and are approved:
- `.sdlc/features/$ARGUMENTS/spec.md`
- `.sdlc/features/$ARGUMENTS/design.md`
- `.sdlc/features/$ARGUMENTS/tasks.md`
If any are missing or not approved, stop and report what needs to be completed first.
### 3. Transition to Implementation (if needed)
If in `ready` phase:
```bash
sdlc feature transition $ARGUMENTS implementation
```
### 4. Get Task List
```bash
sdlc task list $ARGUMENTS --json
```
Identify all pending tasks and their dependency order. Tasks with no `blocked_by` can be worked first.
### 5. Implement Tasks in Order
For each pending task (respecting dependencies):
1. **Start the task:**
```bash
sdlc task start $ARGUMENTS <task-id>
```
2. **Read task scope** from tasks.md -- find the specific task section
3. **Study existing patterns** in files to be modified
4. **Implement changes:**
- Production code first
- Tests alongside or immediately after
- Follow existing conventions
5. **Run tests:**
```bash
go test ./... -v 2>&1 | tee /tmp/task-test-output.txt
# or the appropriate test command for the project stack
```
6. **Complete (only if tests pass):**
```bash
sdlc task complete $ARGUMENTS <task-id>
```
7. **Continue** to next pending task
### 6. Final Verification
After all tasks complete:
```bash
go test ./... -v
go vet ./...
```
### 7. Report Summary
- Tasks completed (count)
- Files changed
- Tests added/modified
- Final test status
## Critical Rules
- ALWAYS verify prerequisites before implementing
- ALWAYS implement tasks in dependency order
- NEVER mark a task complete if tests fail
- NEVER continue to next task if current task fails
- ALWAYS run final verification after all tasks
- ALWAYS follow existing codebase patterns
- NEVER implement beyond task stated scope
- ALWAYS read spec and design for context before coding