File attachments were embedded into AI messages without any PII
scanning, allowing credentials, tokens, and other sensitive data
to be forwarded to AI providers in plaintext.
Typed chat messages had the same gap: a user could type a password
or API key directly and it would be sent unscanned.
Changes:
- chat_message (Rust): defence-in-depth scan of all attachment body
content (between --- Attached: markers); hard rejects if PII found
- detect_pii (Rust): fix return type from pii::PiiDetectionResult
(spans/original_text) to db::models::PiiDetectionResult
(detections/total_pii_found) to match the TypeScript contract; the
LogUpload PII review workflow was receiving undefined for detections
- scan_text_for_pii (Rust): new command — scans arbitrary text for PII
without creating DB records; used for typed message warnings
- Triage/index.tsx: PendingFile now carries logFileId; handleSend gates
each text attachment through detectPiiCmd (hard block on PII found);
typed message text scanned via scanTextForPiiCmd with a one-time
warning — second send of same message proceeds as acknowledgment
AI history continuity: Changed the history-load query in chat_message to
JOIN ai_conversations and select by issue_id instead of single conversation_id.
This preserves full context when provider/model changes mid-triage.
Deep search: Added DISTINCT to list_issues SELECT and extended the search
filter with EXISTS subqueries covering ai_messages, resolution_steps,
log_files, and timeline_events. Ensures comprehensive search without
duplicate results.
Includes 11 new unit tests covering both features.
- Implement AgentRegistry system with devops-incident-responder agent
- Add domain detection based on conversation keywords
- Inject devops-incident-responder as primary system prompt
- Auto-switch domain prompts silently when context shifts
- Fix version update script to handle JSON format correctly
- Always display version in bottom-left corner
- Add release notes fallback to git commits if CHANGELOG empty
This implements the full devops-incident-responder agent as the primary
system prompt, with domain-specific SME prompts layered on top based on
conversation content analysis. The version display bug is fixed by removing
the collapsed condition, and release notes now have a fallback mechanism.
Add INCIDENT_RESPONSE_FRAMEWORK to domainPrompts.ts and append it to
all 17 domain prompts via getDomainPrompt(). Add system_prompt param
to chat_message command so frontend can inject domain expertise. Record
UTC timeline events (triage_started, log_uploaded, why_level_advanced,
root_cause_identified, rca_generated, postmortem_generated,
document_exported) at key moments with non-blocking calls.
Update tauriCommands.ts with getTimelineEventsCmd, optional metadata on
addTimelineEventCmd, and systemPrompt on chatMessageCmd.
12 new frontend tests (9 domain prompts, 3 timeline events).
This commit implements two major features:
1. Integration Search as Primary AI Data Source
- Confluence, ServiceNow, and Azure DevOps searches execute before AI queries
- Search results injected as system context for AI providers
- Parallel search execution for performance
- Webview-based fetch for HttpOnly cookie support
- Persistent browser windows maintain authenticated sessions
2. AI Tool-Calling (Function Calling)
- Allows AI to automatically execute functions during conversation
- Implemented for OpenAI-compatible providers and Custom REST provider
- Created add_ado_comment tool for updating Azure DevOps tickets
- Iterative tool-calling loop supports multi-step workflows
- Extensible architecture for adding new tools
Key Files:
- src-tauri/src/ai/tools.rs (NEW) - Tool definitions
- src-tauri/src/integrations/*_search.rs (NEW) - Integration search modules
- src-tauri/src/integrations/webview_fetch.rs (NEW) - HttpOnly cookie workaround
- src-tauri/src/commands/ai.rs - Tool execution and integration search
- src-tauri/src/ai/openai.rs - Tool-calling for OpenAI and Custom REST provider
- All providers updated with tools parameter support
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
analyze_logs() was reading the original log file from disk and sending its
full contents to external AI providers, completely bypassing the redaction
pipeline. The redacted flag in log_files and the .redacted file on disk were
written by apply_redactions() but never consulted on the read path.
Fix: query the redacted column alongside file_path. If the file has not been
redacted, return an error to the caller before any AI provider call is made.
When redacted, read from {path}.redacted instead of the original.
Adds redacted_path_for() helper and two unit tests covering the rejection
and happy-path cases.
Remove high-risk defaults and tighten data handling across auth, storage, IPC, provider calls, and capabilities so sensitive data is better protected by default. Also update README/wiki security guidance and add targeted tests for the new hardening behaviors.
Made-with: Cursor
Rename custom API format handling from custom_rest to custom_rest with backward compatibility, add guided model selection with custom entry in provider settings, and rebrand app naming to Troubleshooting and RCA Assistant across UI, metadata, and docs.
Made-with: Cursor
Inline format arguments across Rust modules to satisfy clippy -D warnings, and configure Cargo to prefer system OpenSSL so clippy builds do not fail on missing vendored Perl modules.
Made-with: Cursor
- Fix download icons (PDF/DOCX) not visible in dark theme by adding text-foreground class
- Fix "Read-only file system" error by using Downloads directory for exports with proper fallback
- Fix Search button visibility in History page by changing variant and adding icon
- Fix domain-only filtering in History page by adding missing filter.domain handling
- Enhance audit log to capture full transmitted data (provider details, messages, content previews)
- Add dirs crate dependency for cross-platform directory detection
- Add success/error feedback for document exports with file path display
- Update Security page to display pretty-printed JSON audit details
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
