LeoYeAI/agent-master-skills / skills/agent-memory-system-guide
Agent Memory System Guide
VerifiedAn agent memory workflow guide for OpenClaw and Codex with MEMORY.md, daily notes, SESSION-STATE, working-buffer, Obsidian archiving, and optional OpenViking support.
docs/superpowers/plans/2026-03-26-import-clean-mode.md
View on GitHub# Import Clean Mode Implementation Plan
> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
**Goal:** Add an explicit `import --clean` mode that performs a deterministic clean restore of the supported memory surface while preserving the current default import behavior.
**Architecture:** Keep the current import path as the default overwrite-style restore. Extend the import CLI with `--clean`, back it with a helper that removes only supported files and directories after backup creation, and update docs to distinguish conservative import from clean restore.
**Tech Stack:** Python 3, argparse, pathlib, zipfile, pytest, Markdown docs
---
### Task 1: Lock the two import semantics with failing tests
**Files:**
- Modify: `tests/test_memory_capture_script.py`
- Test: `tests/test_memory_capture_script.py`
- [ ] **Step 1: Write the failing tests**
```python
def test_import_preserves_extra_supported_files_by_default(tmp_path: Path):
source_workspace = tmp_path / 'source'
source_memory_dir = source_workspace / 'memory'
source_memory_dir.mkdir(parents=True)
(source_workspace / 'MEMORY.md').write_text('# MEMORY\n\nrestored\n', encoding='utf-8')
(source_memory_dir / '2026-03-25.md').write_text('# daily restored\n', encoding='utf-8')
archive_path = tmp_path / 'memory-backup.zip'
export_result = run_command(
'export',
'--workspace',
str(source_workspace),
'--output',
str(archive_path),
'--generated-at',
'2026-03-26T10:00:00+08:00',
)
assert export_result.returncode == 0
target_workspace = tmp_path / 'target'
target_memory_dir = target_workspace / 'memory'
target_memory_dir.mkdir(parents=True)
(target_workspace / 'MEMORY.md').write_text('# MEMORY\n\nexisting\n', encoding='utf-8')
extra_note = target_memory_dir / '2026-03-20.md'
extra_note.write_text('# extra daily note\n', encoding='utf-8')
result = run_command(
'import',
'--workspace',
str(target_workspace),
'--input',
str(archive_path),
'--generated-at',
'2026-03-26T10:05:00+08:00',
)
assert result.returncode == 0
assert extra_note.exists()
assert 'Import mode: conservative' in result.stdout
def test_import_clean_removes_extra_supported_files_before_restore(tmp_path: Path):
source_workspace = tmp_path / 'source'
source_memory_dir = source_workspace / 'memory'
source_memory_dir.mkdir(parents=True)
(source_workspace / 'MEMORY.md').write_text('# MEMORY\n\nrestored\n', encoding='utf-8')
(source_memory_dir / '2026-03-25.md').write_text('# daily restored\n', encoding='utf-8')
archive_path = tmp_path / 'memory-backup.zip'
export_result = run_command(
'export',
'--workspace',
str(source_workspace),
'--output',
str(archive_path),
'--generated-at',
'2026-03-26T10:10:00+08:00',
)
assert export_result.returncode == 0
target_workspace = tmp_path / 'target'
target_memory_dir = target_workspace / 'memory'
attachments_dir = target_workspace / 'attachments'
target_memory_dir.mkdir(parents=True)
attachments_dir.mkdir()
(target_workspace / 'MEMORY.md').write_text('# MEMORY\n\nexisting\n', encoding='utf-8')
extra_note = target_memory_dir / '2026-03-20.md'
extra_note.write_text('# extra daily note\n', encoding='utf-8')
extra_attachment = attachments_dir / 'old.txt'
extra_attachment.write_text('old\n', encoding='utf-8')
result = run_command(
'import',
'--workspace',
str(target_workspace),
'--input',
str(archive_path),
'--generated-at',
'2026-03-26T10:15:00+08:00',
'--clean',
)
assert result.returncode == 0
assert not extra_note.exists()
assert not extra_attachment.exists()
assert (target_workspace / 'memory' / '2026-03-25.md').exists()
assert 'Import mode: clean' in result.stdout
backups = sorted(target_workspace.glob('memory-import-backup-*.zip'))
assert backups
with zipfile.ZipFile(backups[0]) as archive:
assert archive.read('memory/2026-03-20.md').decode('utf-8') == '# extra daily note\n'
assert archive.read('attachments/old.txt').decode('utf-8') == 'old\n'
```
- [ ] **Step 2: Run tests to verify they fail**
Run: `pytest tests/test_memory_capture_script.py -k "preserves_extra_supported_files_by_default or clean_removes_extra_supported_files_before_restore" -q`
Expected: FAIL because the CLI does not accept `--clean`, does not emit mode lines, and current behavior does not explicitly guarantee the new semantics.
- [ ] **Step 3: Write minimal implementation**
Add the `--clean` flag and deletion helper in `scripts/memory_capture.py`:
```python
import shutil
```
```python
import_parser.add_argument(
"--clean",
action="store_true",
help="Remove supported memory files and directories before restoring the archive.",
)
```
```python
def clear_supported_workspace_state(workspace: Path) -> None:
for name in SUPPORTED_FILES:
candidate = workspace / name
if candidate.exists():
candidate.unlink()
for directory_name in SUPPORTED_DIRECTORIES:
directory = workspace / directory_name
if directory.exists():
shutil.rmtree(directory)
```
```python
with zipfile.ZipFile(archive_path) as archive:
members = safe_members_from_manifest(archive)
backup_path = backup_existing_workspace_state(workspace, generated_at)
if args.clean:
clear_supported_workspace_state(workspace)
restore_archive(archive, workspace, members)
print(f"Import mode: {'clean' if args.clean else 'conservative'}")
```
Keep equivalent final code if helper names differ, but preserve the same behavior and scope.
- [ ] **Step 4: Run tests to verify they pass**
Run: `pytest tests/test_memory_capture_script.py -k "preserves_extra_supported_files_by_default or clean_removes_extra_supported_files_before_restore" -q`
Expected: PASS
- [ ] **Step 5: Commit**
```bash
git add tests/test_memory_capture_script.py scripts/memory_capture.py
git commit -m "feat: add clean import mode"
```
### Task 2: Update user-facing import semantics in docs
**Files:**
- Modify: `README.md`
- Modify: `README_CN.md`
- Modify: `README_EN.md`
- Modify: `SKILL.md`
- [ ] **Step 1: Write the failing contract test**
Add assertions to an existing contract test in `tests/test_memory_capture_contract.py`:
```python
assert '--clean' in readme_text
assert 'conservative' in readme_en_text
assert 'overwrite-style restore' in readme_en_text
assert '保守' in readme_cn_text
assert '--clean' in skill_text
assert 'clean restore' in skill_text
```
- [ ] **Step 2: Run test to verify it fails**
Run: `pytest tests/test_memory_capture_contract.py -q`
Expected: FAIL because the docs do not yet mention `--clean` or the distinction between conservative import and clean restore.
- [ ] **Step 3: Write minimal documentation updates**
Update the import sections in:
- `README.md`
- `README_CN.md`
- `README_EN.md`
- `SKILL.md`
Required wording to introduce:
- Default import is conservative and overwrite-style.
- `--clean` removes supported memory files/directories before restore.
- Pre-import backup still runs in both modes.
- [ ] **Step 4: Run test to verify it passes**
Run: `pytest tests/test_memory_capture_contract.py -q`
Expected: PASS
- [ ] **Step 5: Commit**
```bash
git add README.md README_CN.md README_EN.md SKILL.md tests/test_memory_capture_contract.py
git commit -m "docs: clarify import restore modes"
```
### Task 3: Run regression verification
**Files:**
- Modify: none
- Test: `tests/test_memory_capture_script.py`
- Test: `tests/test_memory_capture_contract.py`
- Test: full `tests/`
- [ ] **Step 1: Run focused regression tests**
Run: `pytest tests/test_memory_capture_script.py tests/test_memory_capture_contract.py -q`
Expected: PASS
- [ ] **Step 2: Run full test suite**
Run: `pytest -q`
Expected: PASS with all repository contract tests green.
- [ ] **Step 3: Inspect scoped diff**
Run: `git diff -- scripts/memory_capture.py tests/test_memory_capture_script.py tests/test_memory_capture_contract.py README.md README_CN.md README_EN.md SKILL.md`
Expected: Only import-clean implementation, tests, and doc wording changes appear.
- [ ] **Step 4: Commit**
```bash
git add scripts/memory_capture.py tests/test_memory_capture_script.py tests/test_memory_capture_contract.py README.md README_CN.md README_EN.md SKILL.md
git commit -m "feat: document deterministic clean restore"
```
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |