Files
claude-mem/tests/user-prompt-tag-stripping.test.ts
T
Alex Newman 2b223b7cd9 feat: Add dual-tag system for meta-observation control (#153)
* feat: Add dual-tag system for meta-observation control

Implements <private> and <claude-mem-context> tag stripping at hook layer
to give users fine-grained control over what gets persisted in observations
and enable future real-time context injection without recursive storage.

**Features:**
- stripMemoryTags() function in save-hook.ts
- Strips both <private> and <claude-mem-context> tags before sending to worker
- Always active (no configuration needed)
- Comprehensive test suite (19 tests, all passing)
- User documentation for <private> tag
- Technical architecture documentation

**Architecture:**
- Edge processing pattern (filter at hook, not worker)
- Defensive type handling with silentDebug
- Supports multiline, nested, and multiple tags
- Enables strategic orchestration for internal tools

**User-Facing:**
- <private> tag for manual privacy control (documented)
- Prevents sensitive data from persisting in observations

**Infrastructure:**
- <claude-mem-context> tag ready for real-time context feature
- Prevents recursive storage when context injection ships

**Files:**
- src/hooks/save-hook.ts: Core implementation
- tests/strip-memory-tags.test.ts: Test suite (19/19 passing)
- docs/public/usage/private-tags.mdx: User guide
- docs/public/docs.json: Navigation update
- docs/context/dual-tag-system-architecture.md: Technical docs
- plugin/scripts/save-hook.js: Built hook

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

* fix: Strip private tags from user prompts and skip memory ops for fully private prompts

Fixes critical privacy bug where <private> tags were not being stripped from
user prompts before storage in user_prompts table, making private content
searchable via mem-search.

Changes:

1. new-hook.ts: Skip memory operations for fully private prompts
   - If cleaned prompt is empty after stripping tags, skip saveUserPrompt
   - Skip worker init to avoid wasting resources on empty prompts
   - Logs: "(fully private - skipped)"

2. save-hook.ts: Skip observations for fully private prompts
   - Check if user prompt was entirely private before creating observations
   - Respects user intent: fully private prompt = no observations at all
   - Prevents "thoughts pop up" issue where private prompts create public observations

3. SessionStore.ts: Add getUserPrompt() method
   - Retrieves prompt text by session_id and prompt_number
   - Used by save-hook to check if prompt was private

4. Tests: Added 4 new tests for fully private prompt detection (16 total, all passing)

5. Docs: Updated private-tags.mdx to reflect correct behavior
   - User prompts ARE now filtered before storage
   - Private content never reaches database or search indices

Privacy Protection:
- Fully private prompts: No user_prompt saved, no worker init, no observations
- Partially private prompts: Tags stripped, content sanitized before storage
- Zero leaks: Private content never indexed or searchable

Addresses reviewer feedback on PR #153 about user prompt filtering.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

* feat: Enhance memory tag handling and indexing in user prompts

- Added a new index `idx_user_prompts_lookup` on `user_prompts` for improved query performance based on `claude_session_id` and `prompt_number`.
- Refactored memory tag stripping functionality into dedicated utility functions: `stripMemoryTagsFromJson` and `stripMemoryTagsFromPrompt` for better separation of concerns and reusability.
- Updated hooks (`new-hook.ts` and `save-hook.ts`) to utilize the new tag stripping functions, ensuring private content is not stored or searchable.
- Removed redundant inline tag stripping functions from hooks to streamline code.
- Added tests for the new tag stripping utilities to ensure functionality and prevent regressions.

---------

Co-authored-by: Claude <noreply@anthropic.com>
2025-11-30 22:57:26 -05:00

141 lines
5.5 KiB
TypeScript

/**
* Integration tests for user prompt tag stripping
* Verifies that <private> and <claude-mem-context> tags are stripped
* from user prompts before storage in the user_prompts table.
*/
import { describe, it } from 'node:test';
import assert from 'node:assert';
import { stripMemoryTagsFromPrompt } from '../dist/utils/tag-stripping.js';
// Alias for clarity in tests (this tests the prompt context version)
const stripMemoryTags = stripMemoryTagsFromPrompt;
describe('User Prompt Tag Stripping', () => {
it('should strip <private> tags from user prompts', () => {
const userPrompt = 'Please analyze this: <private>API_KEY=secret123</private>';
const expected = 'Please analyze this:';
assert.strictEqual(stripMemoryTags(userPrompt), expected);
});
it('should strip <claude-mem-context> tags from user prompts', () => {
const userPrompt = '<claude-mem-context>Past observations...</claude-mem-context> Continue working';
const expected = 'Continue working';
assert.strictEqual(stripMemoryTags(userPrompt), expected);
});
it('should handle prompts with multiple <private> sections', () => {
const userPrompt = '<private>secret1</private> public text <private>secret2</private>';
const expected = 'public text';
assert.strictEqual(stripMemoryTags(userPrompt), expected);
});
it('should handle prompts that are entirely private', () => {
const userPrompt = '<private>This entire prompt should not be stored</private>';
const expected = '';
assert.strictEqual(stripMemoryTags(userPrompt), expected);
});
it('should preserve prompts without tags', () => {
const userPrompt = 'This is a normal prompt without any tags';
const expected = 'This is a normal prompt without any tags';
assert.strictEqual(stripMemoryTags(userPrompt), expected);
});
it('should handle multiline private content in prompts', () => {
const userPrompt = `Before
<private>
Line 1 of secret
Line 2 of secret
Line 3 of secret
</private>
After`;
const expected = 'Before\n\nAfter';
assert.strictEqual(stripMemoryTags(userPrompt), expected);
});
it('should handle mixed tags in user prompts', () => {
const userPrompt = '<claude-mem-context>Context</claude-mem-context> middle <private>private</private> end';
const expected = 'middle end';
assert.strictEqual(stripMemoryTags(userPrompt), expected);
});
it('should handle real-world example: API credentials', () => {
const userPrompt = `<private>
OPENAI_API_KEY=sk-proj-abc123
DATABASE_URL=postgresql://user:pass@host/db
</private>
Please help me connect to this database and run a query`;
const result = stripMemoryTags(userPrompt);
assert.ok(!result.includes('OPENAI_API_KEY'), 'API key should be stripped');
assert.ok(!result.includes('DATABASE_URL'), 'Database URL should be stripped');
assert.ok(!result.includes('<private>'), 'Private tags should be stripped');
assert.ok(result.includes('Please help me connect'), 'Non-private content should remain');
});
it('should handle real-world example: debugging context', () => {
const userPrompt = `I'm getting an error in the authentication flow.
<private>
Internal debugging notes:
- This is for the Smith project
- Deadline is tomorrow
- Using staging environment
</private>
Can you help me fix the token validation?`;
const result = stripMemoryTags(userPrompt);
assert.ok(!result.includes('Smith project'), 'Debug notes should be stripped');
assert.ok(!result.includes('Deadline'), 'Private context should be stripped');
assert.ok(result.includes('authentication flow'), 'Problem description should remain');
assert.ok(result.includes('token validation'), 'Question should remain');
});
it('should handle edge case: only whitespace after tag removal', () => {
const userPrompt = ' <private>everything</private> ';
const expected = '';
assert.strictEqual(stripMemoryTags(userPrompt), expected);
});
it('should handle edge case: unclosed tags (no stripping)', () => {
const userPrompt = 'Text <private>unclosed tag';
const expected = 'Text <private>unclosed tag';
assert.strictEqual(stripMemoryTags(userPrompt), expected);
});
it('should handle non-string input gracefully', () => {
// @ts-expect-error Testing runtime type safety
const result = stripMemoryTags(null);
assert.strictEqual(result, '');
});
// Tests for fully private prompt behavior
it('should return empty string for fully private prompts', () => {
const fullyPrivate = '<private>Everything is private here</private>';
const result = stripMemoryTags(fullyPrivate);
assert.strictEqual(result, '');
});
it('should return empty string for multiple private sections covering entire prompt', () => {
const fullyPrivate = '<private>Part 1</private> <private>Part 2</private> <private>Part 3</private>';
const result = stripMemoryTags(fullyPrivate);
assert.strictEqual(result, '');
});
it('should detect fully private prompts with only whitespace outside tags', () => {
const fullyPrivate = ' <private>Content</private> ';
const result = stripMemoryTags(fullyPrivate);
assert.strictEqual(result, '');
});
it('should not return empty for partially private prompts', () => {
const partiallyPrivate = '<private>Secret</private> Public content here';
const result = stripMemoryTags(partiallyPrivate);
assert.ok(result.trim().length > 0, 'Should have non-empty content');
assert.ok(result.includes('Public'), 'Should contain public content');
});
});