Skip to contentConsole
Ctrl K
Browse documentation
Documentation
Anectico documentation
Connect your agent and your application. Your agent finds the customers a failure affected and follows the evidence to its cause. You open the proof.
01 Start here
Connect your agent and see your first customer story
Connect your agent with one command, send a test customer, ask what happened, and open the proof.
Start the quickstartInstrument an applicationSDK installation, identity, and existing OpenTelemetry.Investigate a customerAsk your agent for a timeline, then open the proof page.Connect an AI agentOne command, a scoped key, and a first read-only question.Fix a setup problemFind the next check for missing data or denied access.
Browse all documentation
Start here6 pages
Instrument21 pages
- Identify customers
- Plan and prove product events
- JavaScript and TypeScript
- Python
- Go
- Feature flags in your application
- Choose a mobile SDK
- iOS
- Android
- React Native
- Flutter
- OpenTelemetry
- Integrate without an SDK
- Record application metrics
- Capture errors and releases
- Record session replay safely
- Record LLM calls
- Capture product events
- Native intake keys
- Runnable integration examples
- Check your instrumentation and correct an identity
Investigate41 pages
- Understand one customer’s story
- Triage an Issue
- Find recently active customers
- Search logs, traces, and metrics
- Explore and compare metrics
- Investigate service health and dashboards
- Investigate with session replay
- Inspect events and customer cohorts
- Investigate accounts and segments
- Investigate agent runs
- Review detected anomalies
- Explain LLM spend per customer
- Check rollout health
- Discover event property schemas
- Manage tracking plans
- Investigate customer problem reports
- Compare affected and successful customers
- Investigate browser experience
- Measure product events
- Manage saved insights
- Put a project on a live screen
- Live screen templates and panels
- Live tail errors and reconnecting
- Watch a saved product metric
- Derive a metric from your logs
- Measure bounded session paths
- Product experiments
- Conclusions: one call, one answer you can repeat
- Send a scheduled report
- Measure web analytics
- Measure revenue
- Run in-product surveys
- Measure heatmaps
- Run read-only SQL over your events
- Compare declared coverage with observations
- Did my fix work?
- Publish a Service Health starter
- Save a finding
- Observed customer request reliability
- Read a support story and record contacts
- Declared account value
Respond9 pages
Manage Anectico16 pages
- The Console
- Manage an organization and projects
- Connect tools and notification delivery
- Review usage, model prices and decision scoring
- Export data
- Decide what recorded content may be used for
- Live viewing links and access
- Ingestion sources and pipelines
- Manage billing and subscriptions
- Export events and audiences to your warehouse
- Erase a person
- Export a person's data
- Erase an account
- See what your agents did
- Manage agent budgets
- Recover a write after a lost reply
AI agents24 pages
- Run an agent end to end
- Connect an AI agent with MCP
- Use the Anectico CLI
- Proof pages
- The result viewer
- Permissions and safe actions
- Scope recipes for agent jobs
- How agent access is enforced
- Revoke an agent's access
- Charts in your conversation
- Evaluate an agent run
- See every agent that is running
- Drive one registered agent
- Replay a real recorded session
- Memory provenance and quarantine
- Record context assembly
- Contain an agent
- Recover an agent connection
- Install the Anectico skills
- Connect Grok
- Connect ChatGPT
- Connect Muse
- Your agent builds on the API
- Wake events and delivery receipts
Build apps10 pages
- Build an app on the REST API
- Pin the contract your app is built on
- Give an app its own key
- Polling, retries and refusals
- Worked example: a team status board
- Worked example: a support page
- Worked example: a Slack on-call bot
- Worked example: a nightly report
- Share a saved aggregate with a browser page
- Publish a saved aggregate on a static page
Reference36 pages
- Configuration
- Authentication and API keys
- Permission scopes
- REST API
- Metrics REST API
- JavaScript SDK API
- Python SDK API
- Go SDK API
- iOS SDK API
- Android SDK API
- React Native SDK API
- Flutter SDK API
- CLI command reference
- MCP tool reference
- Plans, limits, and retention
- MCP client compatibility
- Verify a signed attestation offline
- Agent span and event attributes
- Machine-readable documentation
- Evaluator configuration
- Evaluation results and human judgments
- Evaluation datasets and sandbox replay
- Evaluation experiments
- Evaluation release gates
- Evaluation API and permissions
- Credential revocation and audit behavior
- Connected agent approvals
- Creating project-bound saved resources
- Tracking-plan errors
- What Anectico redacts and withholds
- Integration event verification
- Feature-flag exposure verification
- Integration job verification
- What integration verification proves and what it does not
- Read coverage and intake completeness
- Agent evidence pagination