DISTILL · CORRELATE · REASON · REMEMBER
The Telemetry-Native AI SRE.
Brings its own trusted telemetry, from signal to fix.
Dstl8 continuously distills, correlates, reasons, and remembers across Kubernetes, AWS, OpenTelemetry, Supabase, Vercel, and more. It catches infrastructure, configuration, and code issues with no alert rules to write, and delivers root cause and the fix to Slack, Claude Code, Cursor, and Codex. What your team learns, the platform keeps.
328 incidents resolved · Zero alert rules
13 Kubernetes clusters · Enterprise fintech, first two months.
Works with your stack
Runtime platforms and AI coding tools: connected in minutes. See all integrations →
NEW IN DSTL8 2.0
Every signal. Every layer.
Telemetry Distillation now spans logs, metrics, traces, and events, so Dstl8 covers ground that used to take several tools. Metrics and traces are in preview.
AGENT TO GPU
Follow an AI request from the agent through every model call, including calls routed through LiteLLM, down to the vLLM model server and the NVIDIA or AMD GPU doing the work. Prompts and replies stay out.
APM
Slow endpoints, rising error rates, and new request paths, each tied to the logs, metrics, and deploy behind it, down to the line of code.
Predictive detection
Forecasts on metrics: when a disk will fill, a CPU will hit its limit, or GPUs will max out. On traces, Dstl8 surfaces the worst request paths on its own.
Cross-signal
Logs, metrics, traces, and sentiment read together with Kubernetes events, so every incident is explained by what changed and when.
DISTILL
Catch What Never Alerts.
Telemetry Distillation reads every signal close to the source and extracts sentiment, patterns, anomalies, and severity. Regressions surface the moment a pattern shifts, with zero alert rules to write. Learn more →
CORRELATE
Verify What Agents Ship.
Every signal is tied to Kubernetes topology, your code graph, and deploy events, so root cause is keyed to what changed. Whether an agent or a person shipped it, you know if it works.
Learn more →
REASON
Fix Without Leaving the Agent or Waiting for Ops.
Möbius, our own fine-tuned model, investigates every incident and hands root cause, evidence, and the fix to Slack, Claude Code, Cursor, or Codex. Möbius builds a dashboard around each diagnosis, so the evidence is on screen without maintaining static dashboards. A SKILL.md ships with the MCP server, so your agent knows how to use it.
Learn more →
REMEMBER
Never Solve the Same Incident Twice.
Every investigation and fix lands in a knowledge graph. Before anyone touches a service, the agent surfaces past incidents and prior fixes. Learn more →
// See how Dstl8 closes the loop.
THE LOOP
One Loop. Signal to Fix.
Dstl8 distills, correlates, reasons, and remembers continuously, so every fix makes the next diagnosis faster.
14 days free · No credit card · Full platform access
WHERE THE FIX LANDS
MCP for editors. CLI for agents. Slack for the team.
Root cause and the fix arrive where work already happens, for the engineer, the agent, and the on-call.
MCP SERVER · AI EDITORS
Runtime context inside the editor
Möbius diagnoses, guides, and remembers, with evidence and the fix surfaced in context. No tab switching.
DSTL8 CLI · AGENTS & AUTOMATION
Runtime context for agents and automation
Terminal access plus packaged skills agents invoke directly, so any AI tool, including agentic ones, can use Dstl8 the way you do.
SLACK · TEAMS
Runtime context in the team channel
Incidents, root cause, and the fix posted to the channel your on-call already watches.
Slack
Teams
What is a telemetry-native AI SRE?
Telemetry-Native AI SRE / noun phrase
TELEMETRY-NATIVE
AI
SRE
Why now: Teams ship more code, faster, through more agents. Observability platforms charge to store all of it, and AI SREs built on them only see what was kept. A telemetry-native AI SRE starts close to the source, and AI SREs built on them only see what was kept. A telemetry-native AI SRE reads all of it.
Start Here
Install & Configure Dstl8 in Under 2 Minutes.
Try the Dstl8 CLI and TUI. Install it, run dstl8 setup, and you’re connected in minutes.
brew install control-theory/dstl8/dstl8
dstl8 setupcurl -fsSL https://install.dstl8.ai/script/dstl8-cli | shnpx dstl8nix run github:control-theory/dstl8Download from https://github.com/control-theory/dstl8/releasesQuick Start
# 1. Install the CLI
brew install control-theory/dstl8/dstl8
# 2. Create a Dstl8 account (or `dstl8 login` if you already have one)
dstl8 setup
# 3. Add a source so logs flow in
dstl8 sources add vercel
# 4. Connect your AI agent, auto-detects MCP-compatible clients on your machine and configures them
dstl8 install --all
dstl8 install claude-codeAdd Sources
# Add Sources
dstl8 sources add kubernetes
dstl8 sources add cloudwatch
dstl8 sources add vercel
dstl8 sources add supabase
dstl8 sources add otlp
dstl8 sources add githubSee Dstl8 in Action
Your first root cause, in minutes.
14 days free · No credit card · Full platform access