starmen1265/kb-skills
Archived
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs(02): create phase 2 integration layer plans

Browse Source
This commit is contained in:
starmen1265
2026-03-11 12:29:22 +08:00
parent 0ab884bc2a
commit 637240c546
4 changed files with 844 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
gsd-openclaw/.planning/ROADMAP.md
+8 -3
View File
@@ -45,7 +45,12 @@ Plans:
1. SKILL.md has valid frontmatter (name, description, triggers) and OpenClaw loads the skill successfully
2. Skill triggers on /gsd:* patterns and project management keywords
3. Response formatting adapts to channel capabilities (markdown for rich channels, plain for constrained ones)
**Plans**: TBD
**Plans**: 3 plans
Plans:
- [ ] 02-01-PLAN.md — SKILL.md creation with /gsd:* trigger pattern (OCLW-01, OCLW-02)
- [ ] 02-02-PLAN.md — FeishuResponseFormatter with Card JSON 2.0 support (OCLW-04)
- [ ] 02-03-PLAN.md — Prompt templates for context recovery, command mapping, error recovery (OCLW-02)
### Phase 3: Orchestration
**Goal**: Users can send GSD slash commands through chat and receive captured output with proper error handling
@@ -76,10 +81,10 @@ Phases execute in numeric order: 1 -> 2 -> 3 -> 4
| Phase | Plans Complete | Status | Completed |
|-------|----------------|--------|-----------|
| 1. Core Infrastructure | 3/4 | In Progress | - |
| 2. Integration Layer | 0/TBD | Not started | - |
| 2. Integration Layer | 0/3 | Ready | - |
| 3. Orchestration | 0/TBD | Not started | - |
| 4. GSD Integration | 0/TBD | Not started | - |
---
*Roadmap created: 2026-03-10*
*Last updated: 2026-03-11 - Completed 01-03 tmux session lifecycle*
*Last updated: 2026-03-11 - Planned Phase 2 integration layer*
gsd-openclaw/.planning/phases/02-integration-layer/02-01-PLAN.md
+182
View File
@@ -0,0 +1,182 @@
---
phase: 02-integration-layer
plan: 01
type: execute
wave: 1
depends_on: []
files_modified:
- skill/SKILL.md
- tests/test_skill_md.py
autonomous: true
requirements:
- OCLW-01
- OCLW-02
must_haves:
truths:
- "SKILL.md has valid YAML frontmatter with name and description fields"
- "OpenClaw can load the skill successfully"
- "Skill triggers on /gsd:* command patterns"
artifacts:
- path: "skill/SKILL.md"
provides: "OpenClaw skill definition and trigger patterns"
min_lines: 20
contains: "---"
- path: "tests/test_skill_md.py"
provides: "Tests for SKILL.md validation"
exports: ["TestSkillMd"]
key_links:
- from: "skill/SKILL.md"
to: "OpenClaw Gateway"
via: "skill loading"
pattern: "name:.*description:"
---
<objective>
Create SKILL.md with valid frontmatter that OpenClaw can load and triggers on /gsd:* commands.
Purpose: Define the skill entry point for OpenClaw's skill registry. The SKILL.md file is the "face" of the gsd-openclaw skill that OpenClaw loads to understand when to invoke this skill.
Output:
- skill/SKILL.md - OpenClaw skill definition
- tests/test_skill_md.py - Validation tests
</objective>
<execution_context>
@/home/starmen1265/.claude/get-shit-done/workflows/execute-plan.md
@/home/starmen1265/.claude/get-shit-done/templates/summary.md
</execution_context>
<context>
@.planning/PROJECT.md
@.planning/ROADMAP.md
@.planning/STATE.md
@.planning/phases/02-integration-layer/02-CONTEXT.md
@.planning/phases/02-integration-layer/02-RESEARCH.md
<interfaces>
<!-- Reference: dev-workflow SKILL.md pattern -->
<!-- OpenClaw skills require only `name` and `description` fields -->
<!-- The `description` field serves as the trigger condition -->
Example frontmatter format:
```yaml
---
name: gsd-openclaw
description: GSD project management via /gsd:* commands. Triggers on /gsd:new-project, /gsd:plan, /gsd:status, /gsd:use and other GSD slash commands.
---
```
</interfaces>
</context>
<tasks>
<task type="auto" tdd="true">
<name>Task 1: Create test scaffold for SKILL.md validation</name>
<files>tests/test_skill_md.py</files>
<behavior>
- Test 1: SKILL.md file exists at skill/SKILL.md
- Test 2: SKILL.md has valid YAML frontmatter between --- delimiters
- Test 3: SKILL.md frontmatter contains `name` field
- Test 4: SKILL.md frontmatter contains `description` field
- Test 5: description field contains /gsd: pattern for triggering
</behavior>
<action>
Create tests/test_skill_md.py with test class TestSkillMd containing:
1. test_skill_md_exists():
- Assert skill/SKILL.md file exists
- Use Path(__file__).parent.parent / "skill" / "SKILL.md"
2. test_skill_md_has_valid_frontmatter():
- Read file content
- Assert content starts with "---\n"
- Assert second "---" delimiter exists
- Use yaml.safe_load() to parse frontmatter (import yaml)
3. test_skill_md_has_name_field():
- Parse frontmatter
- Assert "name" key exists
- Assert name is non-empty string
4. test_skill_md_has_description_field():
- Parse frontmatter
- Assert "description" key exists
- Assert description is non-empty string
5. test_skill_md_description_contains_gsd_pattern():
- Parse frontmatter
- Assert "/gsd:" appears in description (for trigger matching)
Follow test pattern from tests/test_state_manager.py (class-based, clear docstrings).
Add yaml to test dependencies if needed (should be available via pytest).
</action>
<verify>
<automated>pytest tests/test_skill_md.py -v</automated>
</verify>
<done>
- tests/test_skill_md.py exists with 5 passing tests
- All tests fail initially (no SKILL.md yet)
</done>
</task>
<task type="auto" tdd="true">
<name>Task 2: Create SKILL.md with valid frontmatter</name>
<files>skill/SKILL.md</files>
<behavior>
- Test: All tests in test_skill_md.py pass
</behavior>
<action>
Create skill/SKILL.md with:
1. YAML frontmatter (between --- delimiters):
- name: gsd-openclaw
- description: GSD project management via /gsd:* commands. Triggers on /gsd:new-project, /gsd:plan, /gsd:status, /gsd:use and other GSD slash commands.
2. Markdown content body (after frontmatter):
- Title: # GSD-OpenClaw Skill
- "## When to Use" section explaining trigger patterns
- "## Commands" table with supported /gsd:* commands
- "## Project Context" section explaining /gsd:use command
User decision (LOCKED): Only /gsd:* command triggers, no natural language.
Commands to document in table:
| Command | Description |
| /gsd:new-project &lt;name&gt; | Create a new GSD project |
| /gsd:use &lt;project&gt; | Switch to a project context |
| /gsd:status | Show current project status |
| /gsd:plan | Run planning phase |
| /gsd:ship | Ship the current phase |
| /gsd:help | Show available commands |
Follow dev-workflow/SKILL.md structure as reference.
</action>
<verify>
<automated>pytest tests/test_skill_md.py -v</automated>
</verify>
<done>
- skill/SKILL.md exists with valid YAML frontmatter
- All 5 tests in test_skill_md.py pass
- SKILL.md has name, description fields with /gsd: trigger pattern
</done>
</task>
</tasks>
<verification>
- Run pytest tests/test_skill_md.py -v (all tests pass)
- Verify SKILL.md can be parsed as valid YAML
- Verify description contains /gsd: pattern for OpenClaw trigger matching
</verification>
<success_criteria>
1. skill/SKILL.md exists with valid YAML frontmatter
2. Frontmatter contains `name: gsd-openclaw`
3. Frontmatter contains `description` with /gsd:* pattern
4. All 5 tests in tests/test_skill_md.py pass
</success_criteria>
<output>
After completion, create `.planning/phases/02-integration-layer/02-01-SUMMARY.md`
</output>
gsd-openclaw/.planning/phases/02-integration-layer/02-02-PLAN.md
+300
View File
@@ -0,0 +1,300 @@
---
phase: 02-integration-layer
plan: 02
type: execute
wave: 1
depends_on: []
files_modified:
- src/gsd_openclaw/response_formatter.py
- tests/test_response_formatter.py
autonomous: true
requirements:
- OCLW-04
must_haves:
truths:
- "Response formatter creates valid Feishu Card JSON 2.0 structure"
- "Success cards have green theme"
- "Error cards have red theme with guidance buttons"
- "Long messages are paginated with [N/M] markers"
- "Cards support interactive buttons with callback values"
artifacts:
- path: "src/gsd_openclaw/response_formatter.py"
provides: "Feishu Card message formatting"
exports: ["FeishuResponseFormatter", "CardTheme", "CardButton"]
- path: "tests/test_response_formatter.py"
provides: "Tests for response formatting"
exports: ["TestFeishuResponseFormatter"]
key_links:
- from: "FeishuResponseFormatter"
to: "Feishu API"
via: "card JSON structure"
pattern: '"schema": "2.0"'
---
<objective>
Create FeishuResponseFormatter class that formats responses as Feishu Card JSON 2.0 messages with three-zone structure (header/content/actions).
Purpose: Provide channel-adaptive response formatting for Feishu. The formatter converts GSD output into structured Feishu Card messages with proper theming, pagination, and interactive elements.
Output:
- src/gsd_openclaw/response_formatter.py - Response formatter implementation
- tests/test_response_formatter.py - Formatter tests
</objective>
<execution_context>
@/home/starmen1265/.claude/get-shit-done/workflows/execute-plan.md
@/home/starmen1265/.claude/get-shit-done/templates/summary.md
</execution_context>
<context>
@.planning/PROJECT.md
@.planning/ROADMAP.md
@.planning/STATE.md
@.planning/phases/02-integration-layer/02-CONTEXT.md
@.planning/phases/02-integration-layer/02-RESEARCH.md
<interfaces>
<!-- Feishu Card JSON 2.0 Structure (from RESEARCH.md) -->
Three-zone layout:
1. header: Title + status color (blue=info, green=success, red=error)
2. elements: Main content area (div, markdown, note components)
3. actions: Interactive buttons (at card bottom)
Card JSON structure:
```json
{
"schema": "2.0",
"config": {"wide_screen_mode": true},
"header": {
"title": {"tag": "plain_text", "content": "..."},
"template": "green"
},
"elements": [
{"tag": "div", "text": {"tag": "lark_md", "content": "..."}}
]
}
```
Button structure:
```json
{
"tag": "action",
"actions": [
{
"tag": "button",
"text": {"tag": "plain_text", "content": "Confirm"},
"type": "primary",
"value": {"action": "approve", "project": "my-auth"}
}
]
}
```
Use `lark_md` tag for markdown content (Feishu's built-in markdown support).
</interfaces>
</context>
<tasks>
<task type="auto" tdd="true">
<name>Task 1: Create test scaffold for FeishuResponseFormatter</name>
<files>tests/test_response_formatter.py</files>
<behavior>
- Test 1: format_success returns valid card with green theme
- Test 2: format_error returns valid card with red theme
- Test 3: format_info returns valid card with blue theme
- Test 4: Card with buttons includes action element
- Test 5: format_paginated adds [N/M] marker
- Test 6: Card has schema 2.0 and wide_screen_mode config
</behavior>
<action>
Create tests/test_response_formatter.py with test class TestFeishuResponseFormatter:
1. test_format_success_returns_valid_card():
- Create FeishuResponseFormatter()
- Call format_success("Test Title", "Test content")
- Assert card["schema"] == "2.0"
- Assert card["header"]["template"] == "green"
- Assert card["header"]["title"]["content"] == "Test Title"
2. test_format_error_returns_red_theme():
- Call format_error("Error", "Something went wrong")
- Assert card["header"]["template"] == "red"
3. test_format_info_returns_blue_theme():
- Call format_info("Info", "Status update")
- Assert card["header"]["template"] == "blue"
4. test_format_with_buttons_includes_actions():
- Create CardButton("OK", {"action": "ok"})
- Call format_success("Test", "Content", buttons=[button])
- Assert elements contains action with tag="action"
- Assert button text is "OK"
5. test_format_paginated_adds_marker():
- Call format_paginated("Title", "Content", page=1, total_pages=3)
- Assert "[1/3]" appears in content
6. test_card_has_required_structure():
- Call format_success("Test", "Content")
- Assert "config" key exists
- Assert config["wide_screen_mode"] == True
- Assert elements[0]["tag"] == "div"
Import from gsd_openclaw.response_formatter: FeishuResponseFormatter, CardTheme, CardButton
Follow test pattern from tests/test_state_manager.py.
</action>
<verify>
<automated>pytest tests/test_response_formatter.py -v</automated>
</verify>
<done>
- tests/test_response_formatter.py exists with 6 tests
- All tests fail initially (no implementation yet)
</done>
</task>
<task type="auto" tdd="true">
<name>Task 2: Implement CardTheme enum and CardButton dataclass</name>
<files>src/gsd_openclaw/response_formatter.py</files>
<behavior>
- Test: CardButton can be created with text, value, button_type
- Test: CardTheme enum has SUCCESS, ERROR, INFO, WARNING values
</behavior>
<action>
Create src/gsd_openclaw/response_formatter.py with:
1. Imports:
- from dataclasses import dataclass
- from typing import Dict, Any, List, Optional
- from enum import Enum
2. CardTheme(Enum):
- SUCCESS = "green"
- ERROR = "red"
- INFO = "blue"
- WARNING = "orange"
3. @dataclass CardButton:
- text: str
- value: Dict[str, Any]
- button_type: str = "primary" # primary, default, danger
Keep implementation minimal - just these data structures.
Tests should now pass for CardTheme and CardButton.
</action>
<verify>
<automated>pytest tests/test_response_formatter.py::TestFeishuResponseFormatter::test_format_success_returns_valid_card -v 2>&1 | head -20 || echo "Expected: tests still fail (formatter not implemented)"</automated>
</verify>
<done>
- CardTheme enum exists with SUCCESS, ERROR, INFO, WARNING
- CardButton dataclass exists with text, value, button_type
</done>
</task>
<task type="auto" tdd="true">
<name>Task 3: Implement FeishuResponseFormatter class</name>
<files>src/gsd_openclaw/response_formatter.py</files>
<behavior>
- Test: All 6 tests in test_response_formatter.py pass
</behavior>
<action>
Add FeishuResponseFormatter class to response_formatter.py:
class FeishuResponseFormatter:
"""Format responses as Feishu Card messages."""
def format_success(self, title: str, content: str,
buttons: Optional[List[CardButton]] = None) -> dict:
"""Format a success response card with green theme."""
return self._build_card(title, content, CardTheme.SUCCESS, buttons)
def format_error(self, title: str, content: str,
buttons: Optional[List[CardButton]] = None) -> dict:
"""Format an error response card with red theme."""
return self._build_card(title, content, CardTheme.ERROR, buttons)
def format_info(self, title: str, content: str,
buttons: Optional[List[CardButton]] = None) -> dict:
"""Format an info response card with blue theme."""
return self._build_card(title, content, CardTheme.INFO, buttons)
def format_paginated(self, title: str, content: str,
page: int, total_pages: int) -> dict:
"""Format a paginated response with [page/total] marker."""
footer = f"\n\n---\n[{page}/{total_pages}]"
return self._build_card(title, content + footer, CardTheme.INFO)
def _build_card(self, title: str, content: str, theme: CardTheme,
buttons: Optional[List[CardButton]] = None) -> dict:
"""Build complete card JSON structure."""
card = {
"schema": "2.0",
"config": {"wide_screen_mode": True},
"header": {
"title": {"tag": "plain_text", "content": title},
"template": theme.value
},
"elements": [
{"tag": "div", "text": {"tag": "lark_md", "content": content}}
]
}
if buttons:
card["elements"].append(self._build_actions(buttons))
return card
def _build_actions(self, buttons: List[CardButton]) -> dict:
"""Build actions element from button list."""
return {
"tag": "action",
"actions": [
{
"tag": "button",
"text": {"tag": "plain_text", "content": btn.text},
"type": btn.button_type,
"value": btn.value
}
for btn in buttons
]
}
User decision (LOCKED):
- Three-zone structure: header + elements + actions
- Error cards use red theme
- Success cards use green theme
- Pagination uses [N/M] format
- Interactive buttons support callback values
All 6 tests should now pass.
</action>
<verify>
<automated>pytest tests/test_response_formatter.py -v</automated>
</verify>
<done>
- FeishuResponseFormatter class implements all format methods
- All 6 tests pass
- Card JSON uses schema 2.0 with three-zone structure
</done>
</task>
</tasks>
<verification>
- Run pytest tests/test_response_formatter.py -v (all tests pass)
- Verify card structure matches Feishu Card JSON 2.0 specification
- Verify theme colors are correct (green=success, red=error, blue=info)
</verification>
<success_criteria>
1. FeishuResponseFormatter class exists in src/gsd_openclaw/response_formatter.py
2. format_success, format_error, format_info, format_paginated methods work
3. Card JSON has schema "2.0" with proper header/elements structure
4. Buttons are rendered as action elements with callback values
5. All 6 tests in tests/test_response_formatter.py pass
</success_criteria>
<output>
After completion, create `.planning/phases/02-integration-layer/02-02-SUMMARY.md`
</output>
gsd-openclaw/.planning/phases/02-integration-layer/02-03-PLAN.md
+354
View File
@@ -0,0 +1,354 @@
---
phase: 02-integration-layer
plan: 03
type: execute
wave: 1
depends_on: []
files_modified:
- src/gsd_openclaw/prompts/__init__.py
- src/gsd_openclaw/prompts/context_recovery.md
- src/gsd_openclaw/prompts/command_mapping.md
- src/gsd_openclaw/prompts/error_recovery.md
- src/gsd_openclaw/prompt_loader.py
- tests/test_prompt_templates.py
autonomous: true
requirements:
- OCLW-02
must_haves:
truths:
- "Prompt templates exist as markdown files with {{variable}} placeholders"
- "PromptLoader can load and render templates with variable substitution"
- "context_recovery.md template helps resume interrupted sessions"
- "command_mapping.md template maps user input to GSD commands"
- "error_recovery.md template generates user-friendly error messages"
artifacts:
- path: "src/gsd_openclaw/prompts/context_recovery.md"
provides: "Session context recovery prompt"
contains: "{{project_name}}"
- path: "src/gsd_openclaw/prompts/command_mapping.md"
provides: "Command mapping prompt"
contains: "{{user_input}}"
- path: "src/gsd_openclaw/prompts/error_recovery.md"
provides: "Error recovery prompt"
contains: "{{error_message}}"
- path: "src/gsd_openclaw/prompt_loader.py"
provides: "Template loading and rendering"
exports: ["PromptLoader"]
- path: "tests/test_prompt_templates.py"
provides: "Tests for prompt templates"
exports: ["TestPromptLoader"]
key_links:
- from: "PromptLoader"
to: "prompts/*.md"
via: "string.Template.safe_substitute"
pattern: "{{variable}}"
---
<objective>
Create prompt templates for context recovery, command mapping, and error recovery, plus a PromptLoader class for loading and rendering templates.
Purpose: Provide reusable prompt templates that Phase 3's orchestrator will use to generate context-aware prompts for Claude. Templates use simple {{variable}} syntax with string.Template for substitution.
Output:
- src/gsd_openclaw/prompts/ directory with 3 template files
- src/gsd_openclaw/prompt_loader.py - Template loader implementation
- tests/test_prompt_templates.py - Template tests
</objective>
<execution_context>
@/home/starmen1265/.claude/get-shit-done/workflows/execute-plan.md
@/home/starmen1265/.claude/get-shit-done/templates/summary.md
</execution_context>
<context>
@.planning/PROJECT.md
@.planning/ROADMAP.md
@.planning/STATE.md
@.planning/phases/02-integration-layer/02-CONTEXT.md
@.planning/phases/02-integration-layer/02-RESEARCH.md
<interfaces>
<!-- Template pattern from RESEARCH.md -->
Use string.Template with safe_substitute() for {{variable}} syntax:
```python
from string import Template
from pathlib import Path
class PromptLoader:
def __init__(self, template_dir: Path):
self.template_dir = template_dir
self._cache: Dict[str, Template] = {}
def render(self, name: str, **variables) -> str:
path = self.template_dir / f"{name}.md"
template = Template(path.read_text(encoding='utf-8'))
return template.safe_substitute(**variables)
```
Note: string.Template uses $variable syntax by default. For {{variable}} syntax,
need to override the delimiter pattern or use simple string substitution.
</interfaces>
</context>
<tasks>
<task type="auto" tdd="true">
<name>Task 1: Create test scaffold for prompt templates</name>
<files>tests/test_prompt_templates.py</files>
<behavior>
- Test 1: context_recovery.md template exists
- Test 2: command_mapping.md template exists
- Test 3: error_recovery.md template exists
- Test 4: PromptLoader.render substitutes variables
- Test 5: PromptLoader handles missing variables gracefully (safe_substitute)
- Test 6: PromptLoader raises error for missing template file
</behavior>
<action>
Create tests/test_prompt_templates.py with test classes:
class TestPromptTemplatesExist:
1. test_context_recovery_template_exists():
- Assert prompts_dir / "context_recovery.md" exists
2. test_command_mapping_template_exists():
- Assert prompts_dir / "command_mapping.md" exists
3. test_error_recovery_template_exists():
- Assert prompts_dir / "error_recovery.md" exists
4. test_templates_contain_variable_placeholders():
- Read each template
- Assert {{variable}} pattern exists in each
class TestPromptLoader:
5. test_render_substitutes_variables():
- Create PromptLoader with prompts directory
- Call render("context_recovery", project_name="test-project", last_active_at="2026-03-11")
- Assert "test-project" appears in output
- Assert "{{project_name}}" does NOT appear
6. test_render_handles_missing_variables():
- Call render with partial variables
- Assert output contains literal {{missing_var}} (safe_substitute behavior)
7. test_render_raises_for_missing_template():
- Call render("nonexistent_template")
- Assert FileNotFoundError is raised
Use Path(__file__).parent.parent / "src" / "gsd_openclaw" / "prompts" for prompts_dir.
Follow test pattern from tests/test_state_manager.py.
</action>
<verify>
<automated>pytest tests/test_prompt_templates.py -v</automated>
</verify>
<done>
- tests/test_prompt_templates.py exists with 7 tests
- Tests fail initially (no templates yet)
</done>
</task>
<task type="auto" tdd="true">
<name>Task 2: Create prompt template files</name>
<files>
- src/gsd_openclaw/prompts/__init__.py
- src/gsd_openclaw/prompts/context_recovery.md
- src/gsd_openclaw/prompts/command_mapping.md
- src/gsd_openclaw/prompts/error_recovery.md
</files>
<behavior>
- Test: All 4 tests in TestPromptTemplatesExist pass
</behavior>
<action>
Create prompts directory and template files:
1. src/gsd_openclaw/prompts/__init__.py:
- Empty file (marks directory as package)
2. src/gsd_openclaw/prompts/context_recovery.md:
```markdown
# Context Recovery
You are resuming a GSD session that was interrupted.
## Session Information
- **Project**: {{project_name}}
- **Last Active**: {{last_active_at}}
- **Current Phase**: {{current_phase}}
- **Commands Executed**: {{command_count}}
## Last Known State
{{last_output_snapshot}}
## Instructions
Continue from where the session left off. Do not repeat completed steps.
```
3. src/gsd_openclaw/prompts/command_mapping.md:
```markdown
# Command Mapping
Map the following user input to a GSD slash command.
## User Input
{{user_input}}
## Available Commands
- /gsd:new-project &lt;name&gt; - Create a new project
- /gsd:use &lt;project&gt; - Switch to a project
- /gsd:status - Show current status
- /gsd:plan - Run planning phase
- /gsd:ship - Ship current phase
- /gsd:help - Show help
## Current Context
- Project: {{current_project}}
- Phase: {{current_phase}}
## Instructions
Determine the most appropriate command based on the user input and context.
```
4. src/gsd_openclaw/prompts/error_recovery.md:
```markdown
# Error Recovery
An error occurred while processing a GSD command.
## Error Details
**Command**: {{command}}
**Error**: {{error_message}}
## Context
- Project: {{project_name}}
- Phase: {{current_phase}}
## Instructions
Generate a user-friendly error message in Chinese that:
1. Explains what went wrong
2. Suggests possible solutions
3. Offers next steps the user can take
```
User decision (LOCKED): Template format is Markdown with {{variable}} syntax.
</action>
<verify>
<automated>pytest tests/test_prompt_templates.py::TestPromptTemplatesExist -v</automated>
</verify>
<done>
- All 3 template files exist with {{variable}} placeholders
- 4 tests in TestPromptTemplatesExist pass
</done>
</task>
<task type="auto" tdd="true">
<name>Task 3: Implement PromptLoader class</name>
<files>src/gsd_openclaw/prompt_loader.py</files>
<behavior>
- Test: All 7 tests in test_prompt_templates.py pass
</behavior>
<action>
Create src/gsd_openclaw/prompt_loader.py:
```python
"""Prompt template loader for GSD-OpenClaw."""
from __future__ import annotations
import re
from pathlib import Path
from typing import Dict, Any
class PromptLoader:
"""Load and render prompt templates with {{variable}} substitution.
Uses simple regex-based substitution for {{variable}} syntax.
Missing variables are left as-is (safe substitution behavior).
"""
def __init__(self, template_dir: Path | None = None):
"""Initialize PromptLoader.
Args:
template_dir: Directory containing template files.
Defaults to src/gsd_openclaw/prompts/.
"""
if template_dir is None:
template_dir = Path(__file__).parent / "prompts"
self.template_dir = Path(template_dir)
self._cache: Dict[str, str] = {}
def get(self, name: str) -> str:
"""Get raw template content by name (cached).
Args:
name: Template name without .md extension.
Returns:
Raw template content.
Raises:
FileNotFoundError: If template file doesn't exist.
"""
if name not in self._cache:
path = self.template_dir / f"{name}.md"
if not path.exists():
raise FileNotFoundError(f"Template not found: {name}")
self._cache[name] = path.read_text(encoding='utf-8')
return self._cache[name]
def render(self, name: str, **variables: Any) -> str:
"""Render a template with variable substitution.
Uses {{variable}} syntax. Missing variables are left as literal
{{variable}} strings (safe substitution).
Args:
name: Template name without .md extension.
**variables: Variable values to substitute.
Returns:
Rendered template content.
Raises:
FileNotFoundError: If template file doesn't exist.
"""
content = self.get(name)
for key, value in variables.items():
content = content.replace("{{" + key + "}}", str(value))
return content
```
Note: Using simple string.replace() for {{variable}} syntax instead of
string.Template which uses $variable syntax. This matches the user's
decision for {{variable}} format.
</action>
<verify>
<automated>pytest tests/test_prompt_templates.py -v</automated>
</verify>
<done>
- PromptLoader class implements get() and render() methods
- All 7 tests pass
- Missing variables are left as {{variable}} (safe substitution)
</done>
</task>
</tasks>
<verification>
- Run pytest tests/test_prompt_templates.py -v (all tests pass)
- Verify all 3 template files have {{variable}} placeholders
- Verify PromptLoader.render substitutes variables correctly
</verification>
<success_criteria>
1. src/gsd_openclaw/prompts/ directory exists with __init__.py
2. context_recovery.md, command_mapping.md, error_recovery.md templates exist
3. PromptLoader class can load and render templates
4. Variable substitution uses {{variable}} syntax
5. All 7 tests in tests/test_prompt_templates.py pass
</success_criteria>
<output>
After completion, create `.planning/phases/02-integration-layer/02-03-SUMMARY.md`
</output>