What AccessFlow can do today, and what's planned
AccessFlow ships as a single open-source product under Apache 2.0 — every capability on this page is in the same release, so the roadmap is a capability map rather than a sales document. The grid below is grouped by what the platform does, not by the release that introduced it; version-by-version milestone scope lives in docs/12-roadmap.md, and the release currently in progress is tracked on the GitHub milestones.
Available now
Everything in this grid is generally available today in v2.7 — self-hosted with Docker Compose or Helm, and governed by the same review pipeline and tamper-evident audit log throughout. The map itself is driven by community input and adapter contributions.
Proxy & data access
- SQL proxy · 6 DBs + custom drivers
- Read-replica routing · result diffing
- Schema explorer & ER view · sample previews
- Datasource health dashboard
- High-volume proxy performance — result caching · query batching · multi-replica load balancing
- Declarative connector catalog
- MongoDB · Couchbase (SQL++) · Redis
- Cassandra & ScyllaDB (CQL)
- Elasticsearch & OpenSearch · DynamoDB (PartiQL)
- Neo4j (Cypher) · text-to-query for NoSQL
- Cloud data warehouses — Snowflake, BigQuery, Databricks SQL (native auth)
- Per-table row limits · bytes-scanned cost caps · per-user data-volume budgets
AI & monitoring
- Risk scoring + index hints · analytics
- OpenAI-compatible & Hugging Face providers
- Editable prompts · Langfuse · text-to-SQL · RAG
- Offline AI fallback — provider fallback pool with local Ollama
- Approval-likelihood prediction — advisory triage signal from your own review history
- Behavioral anomaly detection (UBA)
- AI optimization & index recommendations
- Expanded MCP agent toolset
- Suggested queries — drafts mined from your own approved history, no AI call needed
- In-app help assistant — answers from the shipped documentation, cites its sources, reads no data
- Voyage AI embeddings for RAG and the help assistant
Review & access
- Multi-stage approvals · auto-approve reads
- Bulk decisions · scheduled & recurring runs · plan templates
- Reviewer groups & scoping · JIT access
- Policy-as-code routing · row-level security · masking
- Break-glass emergency access
- Multi-tenant orgs & per-org quotas
- Real-time collaborative editing
- Mobile PWA + one-tap push
- Version history & diff · dry-run sandbox · replay
- Personalized dashboard & weekly digest
- Request chaining & grouping
- Group-based data & API access grants
- Deterministic SQL review rules — eighteen named checks, per-environment severity, lint as you type
- Policy simulator — replay a draft routing, masking or row-security rule against past traffic
- Effective-access explainer — trace why a request would be approved, routed or held, and who can write to a table
- Table, column and query-shape deny-lists · configurable masking strategies
- External policy decision hook — your own policy service can escalate or reject, never approve
- Effective-permission explorer — everything one person can do on a datasource, and where it comes from
API governance
- Outbound API access governance — REST · SOAP · GraphQL · gRPC
- Operation catalog · response masking · text-to-API
- Decision trace for a hypothetical API call — never contacts the governed API
Deployment governance
- Deployment approval governance — gate CI/CD releases behind approval workflows
- Pipelines & environments · per-pipeline trigger grants · AI release-risk analysis
- Fail-closed deployment gate · freeze windows · break-glass · rollback tracking
- GitHub Actions · GitLab CI · Azure Pipelines gate steps
- Version tracking — which release runs on which environment, with environment tags
- Drift reporting — how many releases and days an environment is behind, on demand
- Decision trace for a hypothetical deployment — "would a release at this moment be held?"
- Dashboard widgets for pending reviews, recent deployments and environment drift
- Schema change sets — written once, promoted environment by environment, locked on first promotion
- Opt-in schema drift scan — tables, columns and foreign keys compared, never written
Compliance
- Data classification tagging (PII/PCI/PHI/GDPR)
- Compliance reports & signed exports · Auditor role
- Result-export governance & DLP — watermarked, policy-gated downloads
- Data lifecycle — retention, right-to-erasure & pseudonymization
- Access recertification campaigns
- Privileged-access report — every query admin and break-glass holder, with usage evidence
Auth & audit
- JWT · SAML · OAuth/OIDC · SCIM provisioning · TOTP · RBAC with custom roles
- HMAC-chained audit log + CSV export · the SQL as it actually ran · calling application
- Service accounts for CI and AI agents — API-key only, tool allow-lists, per-identity rate limits
- SIEM audit streaming & WORM archival — Splunk HEC · syslog/CEF · signed HTTPS · S3 Object Lock
Security & ops
- External secrets managers (Vault / AWS / Azure Key Vault)
- Backup / restore & DR tooling — Helm hooks + audit-chain verification
- ServiceNow & Jira ticketing — auto-created tickets + bi-directional status sync
- Automated sensitive-data discovery — scheduled column scanning proposes classification tags, nested document fields included, stale findings retired
- Least-privilege intelligence — unused-grant analytics with advisory revocation recommendations
- Release update hint — a self-hosted install says when a newer stable release exists
- Scheduled job monitor — every background job's cadence, last run, failures and history
Automation & IaC
- Terraform / OpenTofu provider
- Reusable GitHub Actions & GitLab CI template
Deploy
- Docker Compose · Helm 3 · HPA · health probes
How the capability grid maps to releases
AccessFlow releases in small, themed milestones a few weeks apart rather than one large annual version. A capability graduates from the Planned band into a milestone, ships through the normal review-pipeline and audit machinery, and joins the grid above. The grid never says which release introduced an item — docs/12-roadmap.md does, version by version since v1.0.
The two most recent shipped themes: v2.6 answers from what AccessFlow already knows: the query editor offers draft queries mined from the organisation's own approved history, a catalog of named SQL review rules checks every query as it is typed, an effective-access explainer traces why a request would be approved, routed or held, a policy simulator replays a draft rule against past traffic before it is saved, and an in-app help assistant answers questions from the documentation shipped with the running version. v2.7 gives CI pipelines and AI agents service accounts of their own — API-key only, limited to chosen tools, rate-limited, and recording the person they act for; lets an access grant deny tables, columns and query shapes as precisely as it allows them, with per-table row limits, warehouse scan caps and per-person data budgets; hands undecided queries to the organisation's own policy service, which can escalate or reject but never approve; and governs database schema changes written once and promoted environment by environment, reaching production only after they applied cleanly in every environment below it, with an opt-in scan that reports schema drift.
What shipped when, release by release, is in the changelog. The milestone being built now, and every open and closed one before it — with the issues inside them — is on GitHub.
What's next
The Planned band holds capabilities that are not yet scheduled into a milestone; an item leaves it when it lands in a milestone, and enters the grid above when it ships. The absence of an engine here is never a blocker — any JDBC-compatible database can be governed today by uploading its driver JAR, and first-class AccessFlow engines are open to contribution.
Planned
- Native wire-protocol gateway
- Custom analyzer plugins (in-process or over HTTP)
- Aggregate disclosure guard (review for aggregates over classified data)