Calls
–
Tokens saved
–
Avg efficiency
–
CO₂ saved
–
Sessions
–
Avg calls / session
–
Tools used
–
Best single call
–
Avg latency
–
p95 latency
–
Max latency
–
Cost saved
–
Energy saved
–
Tokens processed
–
Decisions stored
–
PII items masked
–
Live monitor
Requests, compression, firewall and policy decisions as they happen, across every Synthelion process on this machine.
Requests
0
Compressions
0
Blocked
0
Awaiting approval
0
Event stream
Connecting…
- Waiting for activity…
AI Compliance
One gate over every Synthelion guard, expressed as policy: risk level, remediation action and the article of law each control answers to.
Engine
–
Controls enabled
–
Transactions
–
Audit trail
–
Controls enabled but not operational
These rules are switched on in the policy, but the guard behind them is disabled — they currently inspect nothing.
Guardrails
Risk level, remediation action and scope per control
| Control | Risk | Action | Scope | On |
|---|
Regulatory documents
Generated from the live configuration and the real audit trail
DPIA and FRIA are pre-filled working drafts: sections needing an organisational judgement are listed as open items for the controller or DPO to complete. PDF output is standard PDF 1.4, not PDF/A.
Traceability: control to legal obligation
A disabled control shows here as an uncovered obligation rather than silently still looking compliant.
| Control | Framework | Article | Obligation | Covered |
|---|
Tokens saved over time
Compression savings by day, selected range
reads ~/.synthelion/savings.jsonl
By tool
Share of tokens saved per tool
updates on refresh
By content type
| Type | Tokens saved |
|---|---|
| Loading… | |
Cleanup
Sessions
one row per synthelion-mcp / CLI process| Session | PID | Calls | Tokens saved | Tools used | First activity | Last activity | |
|---|---|---|---|---|---|---|---|
| Loading… | |||||||
Recent requests
–| Time | Tool | Type | Before | After | Saved | Efficiency | Latency | Session |
|---|---|---|---|---|---|---|---|---|
| Loading… | ||||||||
Cleanup
Recent decisions (session memory)
–- Loading…
Cluster
Master/slave nodes, shared backends and generated deploy files.
This node
Multi-node cluster role — standalone, master, or a slave joined to a master
Node ID
–
Master URL
–
Shared cluster token
••••••••
Join a master
Copy the master's node URL and token from its own Cluster page, then join from here
Joined nodes
a node's own dashboard controls it — this is a fleet view, not remote admin| Node ID | URL | Calls | Tokens saved | Version | Status |
|---|---|---|---|---|---|
| No nodes joined yet. | |||||
Deploy files
Docker/Kubernetes templates for this cluster topology — no token baked in, fill in ${SYNTHELION_NODE_TOKEN} at deploy time
Settings
Defaults for compression, storage backends and diagnostics.
Compression
Default level used when a caller (CLI, MCP tool) doesn't specify one
Project Wiki
Default detail level (1-4) for `synthelion wiki` / the MCP tool when not specified per call
Dashboard
Host/port this web server binds to, and how the browser gets live updates
Requires restarting the dashboard process to take effect — the server is already bound to the current host/port and won't rebind itself.
Cluster storage
Session/analytics store and cross-session memory (RAG) backend
Storage status
read-only, current processSessions on disk
–
Decisions stored
–
Ledger records
–
Active memory backend
–
Privacy
PII detection and masking across prompts, streams and documents.
PrivacyGuard
PII detection & masking across 33 countries, prompt-injection screening — a direct port of Caveman.PrivacyGuard (C#)
Settings
Active on the compress hook by default — every toggle here is adjustable
Requires CPU usage: with this on, Synthelion's own PrivacyGuardML model runs on every analyzed text. It ships inside Synthelion, fully offline, with no third-party model and no download step. OFF by default: the regex + checksum pipeline stays zero-ML, and ML only confirms genuinely sensitive bare values — it never overrides a failed checksum.
When the toggle above is on, the UserPromptSubmit hook rejects the prompt outright (with the full PII/privacy note and AI Act notice as the reason) instead of masking-and-continuing.
Turning "Enabled" off restores exactly the pre-1.2.2 behavior (no privacy pre-pass at all).
Whitelist
Exact-match values PrivacyGuard should never flag or mask (e.g. a public support email)
- No whitelisted values yet.
Live tester
Paste text to see PII detection and prompt-injection screening side by side — nothing here is persisted
PII / Privacy
Score: – · Risk: –
Categories: –
Compliance: –
Masked: –
Prompt injection
Score: – · Risk: –
Categories: –
Proxy
The local gateway that compresses, masks and routes every agent call.
Local privacy/compression proxy
Point an agent's ANTHROPIC_BASE_URL / OPENAI_BASE_URL at this proxy for real, enforced masking + compression — no MCP or hook support required on the agent's side. Off by default; runs independently of the MCP integrations, which keep working exactly the same whether this is on or off.
Status
State: –
Address: –
PID: –
Upstreams
Custom routes
Checked before the built-in Anthropic/OpenAI/Gemini routing, first match wins
| Label | Path prefix | Upstream | |
|---|---|---|---|
| Loading… | |||
One explicit, on-demand request when you click "Fetch list" — never automatic, same policy as the PyPI update check.
Reliability
Failover (max 10 backup upstreams)
Circuit breaker
Rolling-history compression
Once a conversation reaches the threshold, older turns compress at "aggressive" while recent ones stay at the configured default
CCR — reversible compression
Cache originals so an agent can ask for the full detail back via synthelion retrieve --token ...
Response cache
Exact-match only (not semantic/ML) — an identical request within the TTL is served without calling the provider again
Budget & output shaping
Recent calls
Metadata only — prompts/responses are never logged| Time | Method | Path | Upstream | Status | Duration | Saved | Result |
|---|---|---|---|---|---|---|---|
| Loading… | |||||||
Users
Employees, their virtual tokens, subscriptions and live consumption.
Users
–
Active subscriptions
–
Tokens billed
–
Spend
–
Subscription alerts
Enterprise users
Virtual tokens, provider access and plans
| User | Role | Status | Virtual token | Providers | Subscriptions | Actions |
|---|
Consumption by user
Billed tokens and quota headroom
| User | Requests | Tokens | Cost | Quota |
|---|
Most used models
By request count, across all users
- No activity recorded yet.
Provider Keys
Real provider credentials, encrypted and held only by the administrator.
Provider Keys
Manage encrypted provider API keys
| Provider | Label | Upstream URL | Actions |
|---|
Subscriptions
Metered and monthly plans, with live quota and cost per user.
Subscriptions
Manage user subscriptions and quotas
| User | Provider | Type | Model | Tokens Used | Cost USD | Status | Actions |
|---|
Activity
Every proxied call, attributed to a user and a subscription.
Activity Log
Recent proxy requests (metadata only, no prompts)
| Time | User | Provider | Model | Tokens | Cost | Status |
|---|
Add Enterprise User
Add Provider Key
Add Subscription
Security
WAF, EnterpriseGuard and the outbound data-loss firewall.
Firewall & WAF
Request pattern inspection (SQLi, XSS, path traversal, command injection, scanner probes, bad user agents), IP allow/block lists, auto-ban and rate limiting — a direct port of the digitalsolutions WAF, protecting the dashboard's network surface
Settings
Default is detect-only — nothing gets blocked until you turn on Block mode
Rule categories
Auto-ban
Rate limit (firewall)
IP rules
Manual allow/block entries — auto-ban and rate-limit bans also show up here
| IP | Kind | Reason | Expires | |
|---|---|---|---|---|
| No IP rules yet. | ||||
Recent events
Most recent detections/blocks — newest first
| Time | IP | Request | Category | Rule | Severity | Action |
|---|---|---|---|---|---|---|
| No events yet. | ||||||
EnterpriseGuard
Outbound data-loss-prevention firewall — blocks cloud/database/FTP/git credentials, private keys and bulk .env dumps before they reach a model, and lets you define file "security zones" that must never be read into an agent's context (enforced via a PreToolUse hook, synthelion firewall-check). Distinct from PrivacyGuard: this data has no safe redacted form, so it is always block-or-allow, never masked.
Content categories (outbound text scanning):
Live tester
Check a text snippet and/or a file path against EnterpriseGuard — nothing submitted here is stored.
Clients
Per-client protected-path policies, keyed by IP (proxy callers) or MAC (local CLI/hook, via synthelion firewall-check — defaults to this machine's own MAC). A client's paths are additive to the policy above, and only apply while Enabled. A client the proxy sees for the first time is auto-registered here, disabled, labeled "Auto-discovered" — review and enable it to activate its policy.
| Label | IP | MAC | Blocked paths | Enabled | |
|---|---|---|---|---|---|
| No clients registered. | |||||
Recent blocks
Metadata only (category/rule/source/time) — the actual blocked text or path is never logged.
| Time | Source | Category | Rule |
|---|---|---|---|
| No blocks yet. | |||
Doctor
Installation health checks for this machine.
Doctor
- Not run yet.
Version
Installed build and available updates.
Version
Running v—
Only checks PyPI when you click the button — Synthelion makes no network calls otherwise.
Profile
Your dashboard account and credentials.
–
Dashboard administrator
Account
Change the dashboard login
Account information
- Username: –
- Synthelion version: –
- Session store: –
- Memory backend: –
Recent notifications
- Loading…
Notifications
Health, security and subscription signals computed from this machine — never fabricated demo content.
All signals
- Loading…
Local dashboard — reads from ~/.synthelion/. Settings/account changes write to disk.