Last updated: February 2026
- Real-time file watcher capturing edits, creates, deletes with timestamps and user attribution
- Web dashboard (Next.js, React, Tailwind, Framer Motion) with live 2-second auto-refresh
- Full-text instant search across all captured context (file name, user, action type, content)
- SQLite local-first database with Prisma ORM
- CLI companion (
devcontext status,search,watch) - Toast notifications for captured context events
- Keyboard shortcuts (Cmd+K search, Cmd+T timeline toggle)
- Timeline view with grid/chronological toggle
- Expertise mapper (auto-identifies who knows what in the codebase)
- Context synthesis (natural language summaries of recent work)
- Knowledge graph built from actual activity data
- Activity filtering by type, user, and date range
- Enhanced search with grouping, highlighting, and sorting
- Git integration with commit tracking
- LLM-powered context Q&A
- GitHub PR/issue integration with dashboard widgets (PR status, issue linking, CI build status)
- GitLab merge request integration (approvals, unresolved discussions, pipeline tracking)
- CI build status widget with failed-run linkage to recent activity
- Deployment tracking (GitHub/GitLab) with commit/time-window activity linkage
- Context recommendations from activity hotspots and follow-ups
- Deduplicated toast notifications for CI test failures
- Team dashboard with contributor and collaboration hotspots
- User profiles with activity history and focus-file summaries
- Enterprise audit logs with severity filters and integrity hashes
- Role-based access control (RBAC) with role matrix and audit-log permissions
- Field-level at-rest encryption for activity content/metadata/tags with migration support
- GDPR compliance workflows (user export, anonymisation, deletion with dry run)
- Enterprise SSO integration (OIDC) with token validation and role-claim mapping
- PostgreSQL support with dual Prisma schemas and provider-aware runtime client loading
- Cloud deployment (Docker Compose, Vercel, Render, Railway, Fly.io, Kubernetes readiness)
- Multi-workspace support with cookie-based switching
- API rate limiting middleware with per-bucket policies
- VS Code extension
.gitlab-ci.ymlCI/CD pipeline
- One-command installer:
npx create-devcontext/brew install devcontextthat scaffolds the project, initialises the database, starts the watcher, and opens the dashboard — zero manual setup - Guided first-run experience: Dashboard walkthrough that highlights key features (search, timeline, knowledge graph, Q&A) with sample data so users immediately see value
devcontext doctor: Health check command that verifies watcher status, database connectivity, Git hooks, API health, and integration credentials in one pass- Desktop app (Electron/Tauri): Standalone app that bundles the dashboard, watcher, and CLI — no terminal or Node.js required, runs from system tray
- Smart capture filters: Configurable ignore patterns (
.devcontextignore) to exclude build artifacts,node_modules, generated files — reduce noise without losing signal - Session grouping: Auto-detect coding sessions (start/stop based on activity gaps) and label them — "Session: Auth refactor (2h 14m, 47 files)" — instead of a flat activity stream
- Quick-capture notes: Cmd+Shift+N from anywhere to attach a plain-text note to your current context — "Chose Redis over Memcached because of pub/sub needs" — preserving the why behind decisions
- Shareable context snapshots: Export a time-range of context as a self-contained HTML file or markdown summary — hand off to a teammate or use for standup notes
- Dashboard performance optimisation: Virtual scrolling for large activity feeds, incremental search indexing, and lazy-loading for knowledge graph visualisations
- JetBrains IDE plugin: IntelliJ/WebStorm/PyCharm plugin matching VS Code extension capabilities — context capture, inline search, knowledge graph sidebar
- Jira/Linear/Asana integration: Auto-link activity to project management tickets based on branch names, commit messages, or manual tagging
- Slack/Teams bot: Daily digest of your coding context posted to a channel — "Yesterday you touched 12 files across auth and payments, made 3 PRs"
- Calendar integration: Correlate coding sessions with calendar events — "You were working on the payments refactor during the 'Sprint planning' meeting"
- End-to-end encryption for sync: When sharing context between machines or team members, all data encrypted in transit and at rest — zero-knowledge architecture where the server never sees plaintext
- Local-only mode enforcement: Strict mode that guarantees no data leaves the machine — network calls disabled, integrations blocked, full air-gap support
- Sensitive file detection: Auto-detect and redact content from files matching sensitive patterns (
.env,credentials.json,*.pem,*secret*) — prevent accidental capture of secrets - Data retention policies: Configurable auto-deletion of context older than N days/months — comply with data minimisation principles without manual cleanup
- Granular consent controls: Per-integration, per-workspace opt-in — users explicitly approve what data flows where
- SOC 2 Type II readiness: Audit log completeness, access control documentation, and encryption coverage sufficient for SOC 2 certification
- SAML 2.0 SSO: Support for Okta, Azure AD, OneLogin, and generic SAML providers alongside existing OIDC
- IP allowlisting: Restrict dashboard and API access to specific IP ranges for self-hosted deployments
- Session management: Configurable session timeouts, concurrent session limits, and force-logout capabilities
- Vulnerability scanning in CI: Integrate Snyk/Trivy into the DevContext build pipeline itself — practise what you preach
- Cryptographically signed context: Each activity entry signed with a user key — tamper-evident chain proving context hasn't been modified after capture
- Export verification: Recipients of shared context snapshots can verify integrity and authenticity before trusting the data
- Backup encryption: Automatic encrypted backups of the SQLite/PostgreSQL database with scheduled rotation
- "Why does this code exist?" answers: Click any file or function in the knowledge graph and get an AI-generated narrative tracing its history — who created it, what problem it solved, how it evolved, which PRs touched it — the complete archaeological record
- Onboarding autopilot: New team member joins → DevContext generates a personalised onboarding guide based on the team's knowledge graph — "Start with
auth/(Lisa is the expert), thenpayments/(Marcus owns it), skiplegacy/for now" - Predictive context loading: When you open a file, DevContext pre-loads related context (recent changes, who worked on it, linked PRs, known issues) before you even search — context finds you, not the other way around
- "What should I review?" suggestions: Before a PR review, DevContext analyses your expertise profile and the PR's files to highlight what you're uniquely qualified to catch — smart reviewer assignment based on actual knowledge, not CODEOWNERS
- Natural language activity queries: Ask the dashboard questions in plain English — "What did the team work on last week related to payments?", "Who last touched the auth middleware and why?"
- 3D knowledge graph: Interactive three.js-powered knowledge graph where files are nodes, developers are orbiting satellites, and connections glow based on recency — zoom, rotate, click to explore — a visually stunning way to understand your codebase
- "Code DNA" profile: Each developer gets a visual fingerprint showing their expertise distribution, activity patterns, collaboration frequency, and focus areas — beautiful radial charts that make engineering profiles shareable
- Time-travel replay: Scrub through a timeline slider and watch the knowledge graph evolve — see how the codebase grew, who contributed what, and when architectural decisions were made — like git log but visual and alive
- Daily/weekly "context digest" emails: Beautifully formatted email summaries with your top files, collaboration highlights, open threads, and AI-generated "what to focus on today" suggestions
- "Context Score" badge: Embeddable badge for READMEs showing how well-documented a project's context is (
DevContext: 94% Coverage | 3 Active Contributors) — incentivise context capture across teams
- MCP (Model Context Protocol) server: Expose DevContext as an MCP tool so AI coding assistants (Claude, Copilot, Cursor) can query your project's knowledge graph, understand code history, and get context-aware suggestions
- Browser extension: Highlight code on GitHub/GitLab and see DevContext overlays — who last touched it, why it exists, related PRs — context follows you across tools
- Plugin system: Extensible capture adapters — bring your own data sources (Figma activity, Notion edits, terminal commands, browser tabs) and feed them into the DevContext knowledge graph
- Public API with webhooks: REST + WebSocket API for building custom integrations — trigger workflows when context patterns are detected (e.g., "alert when 3+ developers touch the same file in a day")
- DevContext Cloud: Optional hosted sync layer — encrypted cross-machine sync, team dashboards, and centralised knowledge graphs without self-hosting
- Cross-repository knowledge graph: Unified graph spanning all repos in an organisation — see how changes in
shared-utilsripple through downstream projects - "Institutional memory" preservation: When a developer leaves, their knowledge graph remains — the why behind every decision is preserved for the team forever
- AI pair programmer with context: Feed DevContext's knowledge graph into AI coding assistants so they understand not just your code, but your team's conventions, past decisions, and architectural intent
- Meeting context capture: Transcribe and link meeting discussions to code changes — "We decided to use Redis in the Feb 3rd architecture meeting (linked:
cache/redis-client.ts)" - Codebase health scoring: AI-powered assessment of code areas that are under-documented, over-concentrated (bus factor risk), or frequently causing issues — proactive knowledge risk management
- Open protocol for developer context: Publish an open standard for context interchange so tools can share developer context across IDEs, CI systems, and project management platforms
We welcome contributions! See CONTRIBUTING.md for guidelines.
High-impact areas right now:
- 🧠 AI/LLM intelligence improvements (context synthesis, Q&A quality)
- 🔌 New integration adapters (Jira, Linear, Slack, JetBrains)
- 📊 Dashboard visualisation enhancements
- 🔐 Security hardening and privacy features
- 📱 Desktop app (Electron/Tauri) development
- 📚 Documentation and onboarding guides
- 🧪 Test coverage for capture and search edge cases
Report bugs or request features via GitHub Issues.