|
|
@@ -0,0 +1,728 @@
|
|
|
+# Claude Code Plugin - Gap Analysis & Recommendations
|
|
|
+
|
|
|
+**Date**: 2026-02-16
|
|
|
+**Status**: Production-ready with identified gaps
|
|
|
+**PR**: #219
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Executive Summary
|
|
|
+
|
|
|
+The Claude Code plugin is **production-ready** but has **4 critical gaps** that affect user experience and workflow efficiency:
|
|
|
+
|
|
|
+1. ❌ **Missing Planning Command** - No easy way to plan/break down tasks
|
|
|
+2. ❌ **Missing Context Addition Command** - No way to add context files easily
|
|
|
+3. ⚠️ **Inconsistent Subagent Invocation** - Format varies across skills
|
|
|
+4. ⚠️ **Explorer/ContextScout Behavior** - Not fully consistent with OpenCode
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Gap 1: Missing Planning Command ❌
|
|
|
+
|
|
|
+### Current State
|
|
|
+
|
|
|
+**What exists**:
|
|
|
+- `/task-breakdown` skill - Invokes task-manager subagent
|
|
|
+- Requires user to manually invoke via skill syntax
|
|
|
+- Part of Stage 4 (Execute) in 6-stage workflow
|
|
|
+
|
|
|
+**Problem**:
|
|
|
+- Users must type: "Use the task-manager subagent to break down..."
|
|
|
+- No simple slash command like `/oac:plan` or `/oac:breakdown`
|
|
|
+- Planning is buried in execution stage, not easily accessible
|
|
|
+
|
|
|
+### Recommended Solution
|
|
|
+
|
|
|
+**Add new command**: `/oac:plan`
|
|
|
+
|
|
|
+**Purpose**: Make task planning/breakdown easily accessible
|
|
|
+
|
|
|
+**Implementation**:
|
|
|
+```markdown
|
|
|
+---
|
|
|
+name: oac:plan
|
|
|
+description: Plan and break down a complex feature into atomic subtasks
|
|
|
+argument-hint: [feature description]
|
|
|
+---
|
|
|
+
|
|
|
+# Plan Feature
|
|
|
+
|
|
|
+Break down the following feature into atomic subtasks: $ARGUMENTS
|
|
|
+
|
|
|
+## Process
|
|
|
+
|
|
|
+1. **Analyze requirements** - Understand scope and complexity
|
|
|
+2. **Discover context** - Find relevant standards and patterns
|
|
|
+3. **Create task breakdown** - Generate subtask files with dependencies
|
|
|
+4. **Present plan** - Show task structure and execution order
|
|
|
+
|
|
|
+## Invocation
|
|
|
+
|
|
|
+This command invokes the task-manager subagent with pre-loaded context to create:
|
|
|
+- `.tmp/tasks/{feature}/task.json` - Feature metadata
|
|
|
+- `.tmp/tasks/{feature}/subtask_NN.json` - Individual subtasks
|
|
|
+
|
|
|
+## Usage Examples
|
|
|
+
|
|
|
+```bash
|
|
|
+# Plan a feature
|
|
|
+/oac:plan user authentication system
|
|
|
+
|
|
|
+# Plan with specific context
|
|
|
+/oac:plan API rate limiting (security focus)
|
|
|
+
|
|
|
+# Plan with constraints
|
|
|
+/oac:plan payment integration (PCI compliance required)
|
|
|
+```
|
|
|
+
|
|
|
+## Output
|
|
|
+
|
|
|
+Task files created in `.tmp/tasks/{feature}/` with:
|
|
|
+- Clear dependencies
|
|
|
+- Parallel execution flags
|
|
|
+- Acceptance criteria
|
|
|
+- Suggested agents
|
|
|
+```
|
|
|
+
|
|
|
+**Benefits**:
|
|
|
+- ✅ Easy to discover and use
|
|
|
+- ✅ Separates planning from execution
|
|
|
+- ✅ Consistent with other `/oac:*` commands
|
|
|
+- ✅ Reduces cognitive load for users
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Gap 2: Missing Context Addition Command ❌
|
|
|
+
|
|
|
+### Current State
|
|
|
+
|
|
|
+**What exists**:
|
|
|
+- `/oac:setup` - Downloads context from GitHub
|
|
|
+- `/context-manager` skill - Manages context configuration
|
|
|
+- Manual file creation for custom context
|
|
|
+
|
|
|
+**Problem**:
|
|
|
+- No easy way to add custom context files
|
|
|
+- No command to add context from GitHub worktrees
|
|
|
+- Users must manually create files in `.opencode/context/`
|
|
|
+
|
|
|
+### Recommended Solution
|
|
|
+
|
|
|
+**Add new command**: `/oac:add-context`
|
|
|
+
|
|
|
+**Purpose**: Easily add context files from various sources
|
|
|
+
|
|
|
+**Implementation**:
|
|
|
+```markdown
|
|
|
+---
|
|
|
+name: oac:add-context
|
|
|
+description: Add context files from GitHub, worktrees, or local files
|
|
|
+argument-hint: [source] [options]
|
|
|
+---
|
|
|
+
|
|
|
+# Add Context
|
|
|
+
|
|
|
+Add context files to your project: $ARGUMENTS
|
|
|
+
|
|
|
+## Sources
|
|
|
+
|
|
|
+### 1. GitHub Repository
|
|
|
+```bash
|
|
|
+# Add from GitHub repo
|
|
|
+/oac:add-context github:owner/repo
|
|
|
+
|
|
|
+# Add specific path
|
|
|
+/oac:add-context github:owner/repo/path/to/context
|
|
|
+
|
|
|
+# Add specific branch/tag
|
|
|
+/oac:add-context github:owner/repo#branch-name
|
|
|
+```
|
|
|
+
|
|
|
+### 2. Git Worktree
|
|
|
+```bash
|
|
|
+# Add from worktree
|
|
|
+/oac:add-context worktree:/path/to/worktree
|
|
|
+
|
|
|
+# Add specific subdirectory
|
|
|
+/oac:add-context worktree:/path/to/worktree/context
|
|
|
+```
|
|
|
+
|
|
|
+### 3. Local Files
|
|
|
+```bash
|
|
|
+# Add local file
|
|
|
+/oac:add-context file:./path/to/context.md
|
|
|
+
|
|
|
+# Add local directory
|
|
|
+/oac:add-context file:./path/to/context/
|
|
|
+```
|
|
|
+
|
|
|
+### 4. URL
|
|
|
+```bash
|
|
|
+# Add from URL
|
|
|
+/oac:add-context url:https://example.com/context.md
|
|
|
+```
|
|
|
+
|
|
|
+## Options
|
|
|
+
|
|
|
+- `--category=<name>` - Specify context category (default: custom)
|
|
|
+- `--priority=<level>` - Set priority (critical, high, medium)
|
|
|
+- `--overwrite` - Overwrite existing files
|
|
|
+- `--dry-run` - Preview what would be added
|
|
|
+
|
|
|
+## Examples
|
|
|
+
|
|
|
+```bash
|
|
|
+# Add team standards from GitHub
|
|
|
+/oac:add-context github:acme-corp/standards --category=team
|
|
|
+
|
|
|
+# Add from worktree
|
|
|
+/oac:add-context worktree:../team-context --category=team
|
|
|
+
|
|
|
+# Add custom pattern
|
|
|
+/oac:add-context file:./docs/patterns/auth.md --category=custom --priority=high
|
|
|
+
|
|
|
+# Preview before adding
|
|
|
+/oac:add-context github:owner/repo --dry-run
|
|
|
+```
|
|
|
+
|
|
|
+## What It Does
|
|
|
+
|
|
|
+1. **Fetches content** from specified source
|
|
|
+2. **Validates format** (checks for proper markdown structure)
|
|
|
+3. **Copies to** `.opencode/context/{category}/`
|
|
|
+4. **Updates navigation** (adds to navigation.md)
|
|
|
+5. **Verifies** context is discoverable
|
|
|
+
|
|
|
+## Output
|
|
|
+
|
|
|
+```
|
|
|
+✅ Added 3 context files to .opencode/context/team/
|
|
|
+
|
|
|
+Files added:
|
|
|
+- .opencode/context/team/standards/code-quality.md
|
|
|
+- .opencode/context/team/patterns/auth.md
|
|
|
+- .opencode/context/team/workflows/deployment.md
|
|
|
+
|
|
|
+Updated navigation:
|
|
|
+- .opencode/context/team/navigation.md
|
|
|
+
|
|
|
+Verification:
|
|
|
+✅ All files discoverable via /context-discovery
|
|
|
+```
|
|
|
+```
|
|
|
+
|
|
|
+**Benefits**:
|
|
|
+- ✅ Easy context addition from multiple sources
|
|
|
+- ✅ Supports GitHub worktrees (key requirement)
|
|
|
+- ✅ Validates and organizes automatically
|
|
|
+- ✅ Updates navigation for discoverability
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Gap 3: Inconsistent Subagent Invocation ⚠️
|
|
|
+
|
|
|
+### Current State
|
|
|
+
|
|
|
+**Skill invocation format** (via frontmatter):
|
|
|
+```markdown
|
|
|
+---
|
|
|
+name: context-discovery
|
|
|
+context: fork
|
|
|
+agent: context-scout
|
|
|
+---
|
|
|
+```
|
|
|
+
|
|
|
+**Problem**: Skills pass context inconsistently to subagents
|
|
|
+
|
|
|
+### Analysis of Current Patterns
|
|
|
+
|
|
|
+#### Pattern 1: Context Discovery (GOOD)
|
|
|
+```markdown
|
|
|
+# File: skills/context-discovery/SKILL.md
|
|
|
+
|
|
|
+---
|
|
|
+context: fork
|
|
|
+agent: context-scout
|
|
|
+---
|
|
|
+
|
|
|
+Discover context files for: **$ARGUMENTS**
|
|
|
+```
|
|
|
+
|
|
|
+**What happens**:
|
|
|
+1. Main agent invokes skill
|
|
|
+2. Skill forks to context-scout subagent
|
|
|
+3. Subagent receives: "Discover context files for: {user request}"
|
|
|
+4. ✅ Clean, simple, works well
|
|
|
+
|
|
|
+#### Pattern 2: Task Breakdown (GOOD)
|
|
|
+```markdown
|
|
|
+# File: skills/task-breakdown/SKILL.md
|
|
|
+
|
|
|
+---
|
|
|
+context: fork
|
|
|
+agent: task-manager
|
|
|
+---
|
|
|
+
|
|
|
+Break down this feature into atomic subtasks: $ARGUMENTS
|
|
|
+
|
|
|
+## Your Task
|
|
|
+{detailed instructions}
|
|
|
+```
|
|
|
+
|
|
|
+**What happens**:
|
|
|
+1. Main agent invokes skill
|
|
|
+2. Skill forks to task-manager subagent
|
|
|
+3. Subagent receives full skill content as prompt
|
|
|
+4. ✅ Includes instructions and context
|
|
|
+
|
|
|
+#### Pattern 3: Code Execution (NEEDS IMPROVEMENT)
|
|
|
+```markdown
|
|
|
+# File: skills/code-execution/SKILL.md
|
|
|
+
|
|
|
+---
|
|
|
+context: fork
|
|
|
+agent: coder-agent
|
|
|
+---
|
|
|
+
|
|
|
+# Code Execution Skill
|
|
|
+
|
|
|
+Implement: $ARGUMENTS
|
|
|
+
|
|
|
+{long instructions about what to do}
|
|
|
+```
|
|
|
+
|
|
|
+**Problem**:
|
|
|
+- Instructions are in skill file, not clear if subagent sees them
|
|
|
+- No explicit "Context to load first" section
|
|
|
+- Unclear how pre-loaded context is passed
|
|
|
+
|
|
|
+### Recommended Solution
|
|
|
+
|
|
|
+**Standardize all skill → subagent invocations**:
|
|
|
+
|
|
|
+```markdown
|
|
|
+---
|
|
|
+name: skill-name
|
|
|
+description: Brief description
|
|
|
+context: fork
|
|
|
+agent: subagent-name
|
|
|
+---
|
|
|
+
|
|
|
+# Skill Name
|
|
|
+
|
|
|
+> **Subagent**: {subagent-name}
|
|
|
+> **Purpose**: {what this does}
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Task
|
|
|
+
|
|
|
+{Clear, concise task description}: $ARGUMENTS
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Context Pre-Loaded
|
|
|
+
|
|
|
+The main agent has already loaded these context files (Stage 3):
|
|
|
+
|
|
|
+- {list of context files from Stage 1 discovery}
|
|
|
+
|
|
|
+**Important**: Do NOT attempt to discover context. Use what's provided above.
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Instructions
|
|
|
+
|
|
|
+{Step-by-step instructions for subagent}
|
|
|
+
|
|
|
+1. {Step 1}
|
|
|
+2. {Step 2}
|
|
|
+3. {Step 3}
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Deliverables
|
|
|
+
|
|
|
+Return:
|
|
|
+- {Expected output 1}
|
|
|
+- {Expected output 2}
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Quality Checklist
|
|
|
+
|
|
|
+Before returning, verify:
|
|
|
+- [ ] {Criterion 1}
|
|
|
+- [ ] {Criterion 2}
|
|
|
+```
|
|
|
+
|
|
|
+**Benefits**:
|
|
|
+- ✅ Consistent format across all skills
|
|
|
+- ✅ Explicit context passing
|
|
|
+- ✅ Clear instructions for subagents
|
|
|
+- ✅ Quality checklist ensures completeness
|
|
|
+
|
|
|
+### Files to Update
|
|
|
+
|
|
|
+1. `skills/code-execution/SKILL.md` - Add context section
|
|
|
+2. `skills/test-generation/SKILL.md` - Add context section
|
|
|
+3. `skills/code-review/SKILL.md` - Add context section
|
|
|
+4. `skills/external-scout/SKILL.md` - Clarify context handling
|
|
|
+5. `skills/parallel-execution/SKILL.md` - Add context section
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Gap 4: Explorer/ContextScout Behavior ⚠️
|
|
|
+
|
|
|
+### Current State
|
|
|
+
|
|
|
+**ContextScout subagent**:
|
|
|
+- ✅ Navigation-driven discovery (reads navigation.md)
|
|
|
+- ✅ Read-only (Glob, Grep, Read tools only)
|
|
|
+- ✅ Returns ranked results (Critical → High → Medium)
|
|
|
+- ✅ Verifies paths exist before recommending
|
|
|
+
|
|
|
+**Comparison with OpenCode**:
|
|
|
+
|
|
|
+| Feature | OpenCode | Claude Code | Status |
|
|
|
+|---------|----------|-------------|--------|
|
|
|
+| Navigation-driven | ✅ | ✅ | ✅ Same |
|
|
|
+| Read-only tools | ✅ | ✅ | ✅ Same |
|
|
|
+| Ranked results | ✅ | ✅ | ✅ Same |
|
|
|
+| Global fallback | ✅ | ✅ | ✅ Same |
|
|
|
+| Nested calls | ✅ | ❌ | ⚠️ Different (by design) |
|
|
|
+| Context pre-loading | ❌ | ✅ | ⚠️ Different (required) |
|
|
|
+
|
|
|
+### Key Differences (By Design)
|
|
|
+
|
|
|
+#### 1. Nested Calls
|
|
|
+
|
|
|
+**OpenCode**:
|
|
|
+```
|
|
|
+Main Agent → TaskManager → CoderAgent → ContextScout
|
|
|
+```
|
|
|
+- Subagents can call ContextScout when needed
|
|
|
+- Dynamic context discovery during execution
|
|
|
+
|
|
|
+**Claude Code**:
|
|
|
+```
|
|
|
+Main Agent → ContextScout (Stage 1)
|
|
|
+Main Agent → Read all files (Stage 3)
|
|
|
+Main Agent → TaskManager/CoderAgent (Stage 4)
|
|
|
+```
|
|
|
+- Only main agent can call ContextScout
|
|
|
+- Context pre-loaded before execution
|
|
|
+- Prevents nested calls (Claude Code constraint)
|
|
|
+
|
|
|
+**Impact**: ✅ Acceptable - enforced by Claude Code architecture
|
|
|
+
|
|
|
+#### 2. Context Pre-Loading
|
|
|
+
|
|
|
+**OpenCode**:
|
|
|
+- Subagents discover context as needed
|
|
|
+- Lazy loading during execution
|
|
|
+
|
|
|
+**Claude Code**:
|
|
|
+- Main agent discovers all context upfront (Stage 1)
|
|
|
+- Main agent loads all context (Stage 3)
|
|
|
+- Subagents use pre-loaded context (Stage 4+)
|
|
|
+
|
|
|
+**Impact**: ✅ Acceptable - required for flat hierarchy
|
|
|
+
|
|
|
+### Recommended Improvements
|
|
|
+
|
|
|
+#### Improvement 1: Explicit Context Passing
|
|
|
+
|
|
|
+**Current**: Subagents assume context is available
|
|
|
+
|
|
|
+**Recommended**: Main agent explicitly passes context in delegation
|
|
|
+
|
|
|
+**Example**:
|
|
|
+```markdown
|
|
|
+Invoke coder-agent subagent:
|
|
|
+
|
|
|
+**Task**: Implement JWT authentication service
|
|
|
+
|
|
|
+**Context Pre-Loaded** (from Stage 3):
|
|
|
+- .opencode/context/core/standards/code-quality.md
|
|
|
+- .opencode/context/core/standards/security-patterns.md
|
|
|
+- .opencode/context/core/standards/typescript.md
|
|
|
+
|
|
|
+**Key Requirements** (extracted from context):
|
|
|
+- Use functional patterns (no classes)
|
|
|
+- RS256 algorithm for JWT signing
|
|
|
+- Token expiry: 15 minutes (access), 7 days (refresh)
|
|
|
+- Error handling: throw typed errors
|
|
|
+
|
|
|
+**Instructions**: Implement following loaded standards.
|
|
|
+```
|
|
|
+
|
|
|
+**Benefits**:
|
|
|
+- ✅ Subagent knows exactly what context is available
|
|
|
+- ✅ Key requirements extracted and highlighted
|
|
|
+- ✅ No ambiguity about what to follow
|
|
|
+
|
|
|
+#### Improvement 2: ContextScout Recommendations
|
|
|
+
|
|
|
+**Current**: ContextScout returns file paths only
|
|
|
+
|
|
|
+**Recommended**: ContextScout also extracts key requirements
|
|
|
+
|
|
|
+**Example**:
|
|
|
+```markdown
|
|
|
+# Context Files Found
|
|
|
+
|
|
|
+## Critical Priority
|
|
|
+
|
|
|
+**File**: `.opencode/context/core/standards/code-quality.md`
|
|
|
+**Contains**: Code quality standards, functional patterns, error handling
|
|
|
+**Why**: Defines coding patterns you must follow
|
|
|
+
|
|
|
+**Key Requirements**:
|
|
|
+- Functional patterns (no classes)
|
|
|
+- Pure functions where possible
|
|
|
+- Explicit error handling
|
|
|
+- TypeScript strict mode
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+**File**: `.opencode/context/core/standards/security-patterns.md`
|
|
|
+**Contains**: Security best practices, auth patterns
|
|
|
+**Why**: Critical for authentication implementation
|
|
|
+
|
|
|
+**Key Requirements**:
|
|
|
+- RS256 for JWT signing
|
|
|
+- Token rotation required
|
|
|
+- Secure secret storage
|
|
|
+- Rate limiting on auth endpoints
|
|
|
+```
|
|
|
+
|
|
|
+**Benefits**:
|
|
|
+- ✅ Main agent can extract requirements immediately
|
|
|
+- ✅ Reduces need to re-read files
|
|
|
+- ✅ Highlights critical requirements
|
|
|
+- ✅ Faster context application
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Recommendations Summary
|
|
|
+
|
|
|
+### High Priority (Implement Before Merge)
|
|
|
+
|
|
|
+1. **Add `/oac:plan` command** - Makes planning easily accessible
|
|
|
+2. **Add `/oac:add-context` command** - Enables GitHub worktree integration
|
|
|
+3. **Standardize skill → subagent format** - Ensures consistent context passing
|
|
|
+
|
|
|
+### Medium Priority (Post-Merge)
|
|
|
+
|
|
|
+4. **Enhance ContextScout output** - Extract key requirements from context files
|
|
|
+5. **Add context validation** - Verify context files are properly formatted
|
|
|
+6. **Add context versioning** - Track context file versions and updates
|
|
|
+
|
|
|
+### Low Priority (Future Enhancement)
|
|
|
+
|
|
|
+7. **Add context search** - Search across all context files
|
|
|
+8. **Add context templates** - Pre-built context file templates
|
|
|
+9. **Add context analytics** - Track which context files are most used
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Implementation Plan
|
|
|
+
|
|
|
+### Phase 1: Critical Commands (1-2 hours)
|
|
|
+
|
|
|
+**Task 1**: Create `/oac:plan` command
|
|
|
+- File: `plugins/claude-code/commands/oac-plan.md`
|
|
|
+- Invokes: task-manager subagent
|
|
|
+- Output: Task breakdown in `.tmp/tasks/{feature}/`
|
|
|
+
|
|
|
+**Task 2**: Create `/oac:add-context` command
|
|
|
+- File: `plugins/claude-code/commands/oac-add-context.md`
|
|
|
+- Supports: GitHub, worktrees, local files, URLs
|
|
|
+- Output: Context files in `.opencode/context/{category}/`
|
|
|
+
|
|
|
+**Task 3**: Update help documentation
|
|
|
+- File: `plugins/claude-code/commands/oac-help.md`
|
|
|
+- Add: New commands to command list
|
|
|
+- Add: Usage examples
|
|
|
+
|
|
|
+### Phase 2: Standardize Skills (2-3 hours)
|
|
|
+
|
|
|
+**Task 1**: Create skill template
|
|
|
+- File: `plugins/claude-code/context/openagents-repo/templates/skill-template.md`
|
|
|
+- Include: Standard sections (Task, Context, Instructions, Deliverables)
|
|
|
+
|
|
|
+**Task 2**: Update existing skills
|
|
|
+- Files: All `skills/*/SKILL.md` files
|
|
|
+- Apply: Standard format
|
|
|
+- Add: Explicit context passing sections
|
|
|
+
|
|
|
+**Task 3**: Update subagent documentation
|
|
|
+- Files: All `agents/*.md` files
|
|
|
+- Clarify: Context pre-loading expectations
|
|
|
+- Add: Examples of context usage
|
|
|
+
|
|
|
+### Phase 3: Enhance ContextScout (3-4 hours)
|
|
|
+
|
|
|
+**Task 1**: Update ContextScout prompt
|
|
|
+- File: `plugins/claude-code/agents/context-scout.md`
|
|
|
+- Add: Key requirement extraction
|
|
|
+- Add: Requirement prioritization
|
|
|
+
|
|
|
+**Task 2**: Update context-discovery skill
|
|
|
+- File: `plugins/claude-code/skills/context-discovery/SKILL.md`
|
|
|
+- Update: Response format to include requirements
|
|
|
+- Add: Examples of enhanced output
|
|
|
+
|
|
|
+**Task 3**: Update using-oac workflow
|
|
|
+- File: `plugins/claude-code/skills/using-oac/SKILL.md`
|
|
|
+- Update: Stage 3 to extract requirements
|
|
|
+- Add: Requirement passing to subagents
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Testing Plan
|
|
|
+
|
|
|
+### Test 1: Planning Command
|
|
|
+
|
|
|
+**Scenario**: User wants to plan a complex feature
|
|
|
+
|
|
|
+**Steps**:
|
|
|
+1. Run: `/oac:plan user authentication system`
|
|
|
+2. Verify: Task files created in `.tmp/tasks/user-authentication-system/`
|
|
|
+3. Verify: Subtasks have dependencies and parallel flags
|
|
|
+4. Verify: Context files referenced correctly
|
|
|
+
|
|
|
+**Expected**: Task breakdown with 4-6 subtasks, clear dependencies
|
|
|
+
|
|
|
+### Test 2: Add Context Command
|
|
|
+
|
|
|
+**Scenario**: User wants to add team standards from GitHub
|
|
|
+
|
|
|
+**Steps**:
|
|
|
+1. Run: `/oac:add-context github:acme-corp/standards --category=team`
|
|
|
+2. Verify: Files downloaded to `.opencode/context/team/`
|
|
|
+3. Verify: Navigation updated
|
|
|
+4. Run: `/context-discovery team coding standards`
|
|
|
+5. Verify: New context files discovered
|
|
|
+
|
|
|
+**Expected**: Context files added and discoverable
|
|
|
+
|
|
|
+### Test 3: Consistent Invocation
|
|
|
+
|
|
|
+**Scenario**: User implements a feature using multiple subagents
|
|
|
+
|
|
|
+**Steps**:
|
|
|
+1. Run: `/context-discovery authentication`
|
|
|
+2. Verify: Context files returned with requirements
|
|
|
+3. Run: `/task-breakdown authentication system`
|
|
|
+4. Verify: Task manager receives context list
|
|
|
+5. Run: `/code-execution implement JWT service`
|
|
|
+6. Verify: Coder agent receives context and requirements
|
|
|
+
|
|
|
+**Expected**: All subagents receive consistent context format
|
|
|
+
|
|
|
+### Test 4: ContextScout Enhancement
|
|
|
+
|
|
|
+**Scenario**: User discovers context for security-sensitive feature
|
|
|
+
|
|
|
+**Steps**:
|
|
|
+1. Run: `/context-discovery payment processing`
|
|
|
+2. Verify: Context files returned
|
|
|
+3. Verify: Key requirements extracted (PCI compliance, encryption, etc.)
|
|
|
+4. Verify: Requirements prioritized
|
|
|
+
|
|
|
+**Expected**: Context files + extracted requirements
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Risk Assessment
|
|
|
+
|
|
|
+### Risk 1: Breaking Changes
|
|
|
+
|
|
|
+**Risk**: Updating skill format breaks existing workflows
|
|
|
+
|
|
|
+**Mitigation**:
|
|
|
+- Maintain backward compatibility
|
|
|
+- Add new sections without removing old ones
|
|
|
+- Test all existing workflows after changes
|
|
|
+
|
|
|
+**Likelihood**: Low
|
|
|
+**Impact**: High
|
|
|
+**Mitigation Status**: ✅ Planned
|
|
|
+
|
|
|
+### Risk 2: Command Confusion
|
|
|
+
|
|
|
+**Risk**: Too many commands confuse users
|
|
|
+
|
|
|
+**Mitigation**:
|
|
|
+- Clear naming conventions (`/oac:*`)
|
|
|
+- Comprehensive help documentation
|
|
|
+- Examples in `/oac:help`
|
|
|
+
|
|
|
+**Likelihood**: Medium
|
|
|
+**Impact**: Low
|
|
|
+**Mitigation Status**: ✅ Planned
|
|
|
+
|
|
|
+### Risk 3: Context Overload
|
|
|
+
|
|
|
+**Risk**: Adding too much context slows down workflow
|
|
|
+
|
|
|
+**Mitigation**:
|
|
|
+- Lazy loading (only load what's needed)
|
|
|
+- Prioritization (Critical → High → Medium)
|
|
|
+- Caching (avoid re-downloading)
|
|
|
+
|
|
|
+**Likelihood**: Low
|
|
|
+**Impact**: Medium
|
|
|
+**Mitigation Status**: ✅ Already implemented
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Success Criteria
|
|
|
+
|
|
|
+### User Experience
|
|
|
+
|
|
|
+- ✅ Users can plan features with one command (`/oac:plan`)
|
|
|
+- ✅ Users can add context from GitHub/worktrees easily
|
|
|
+- ✅ Users understand what context is loaded and why
|
|
|
+- ✅ Subagents receive consistent, clear instructions
|
|
|
+
|
|
|
+### Technical Quality
|
|
|
+
|
|
|
+- ✅ All skills follow standard format
|
|
|
+- ✅ Context passing is explicit and verifiable
|
|
|
+- ✅ ContextScout extracts key requirements
|
|
|
+- ✅ No breaking changes to existing workflows
|
|
|
+
|
|
|
+### Documentation
|
|
|
+
|
|
|
+- ✅ All new commands documented in `/oac:help`
|
|
|
+- ✅ Examples provided for each command
|
|
|
+- ✅ Skill template available for reference
|
|
|
+- ✅ Migration guide for existing users
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+## Conclusion
|
|
|
+
|
|
|
+The Claude Code plugin is **production-ready** with the current feature set, but adding these enhancements will significantly improve user experience and workflow efficiency.
|
|
|
+
|
|
|
+**Recommendation**:
|
|
|
+1. ✅ Merge PR #219 as-is (current state is functional)
|
|
|
+2. 🔄 Create follow-up PR for Phase 1 (critical commands)
|
|
|
+3. 🔄 Create follow-up PR for Phase 2 (standardization)
|
|
|
+4. 🔄 Create follow-up PR for Phase 3 (enhancements)
|
|
|
+
|
|
|
+**Timeline**:
|
|
|
+- Phase 1: 1-2 hours (high priority)
|
|
|
+- Phase 2: 2-3 hours (medium priority)
|
|
|
+- Phase 3: 3-4 hours (low priority)
|
|
|
+
|
|
|
+**Total effort**: 6-9 hours across 3 PRs
|
|
|
+
|
|
|
+---
|
|
|
+
|
|
|
+**Last Updated**: 2026-02-16
|
|
|
+**Author**: OpenAgents Control Team
|
|
|
+**Status**: Ready for Review
|