Skene
Product
Pricing
Docs
Blog
About
Log In
ProductPricingDocsBlogAbout
Log In
Resources / Glossary

Coding agent

An AI tool that reads and writes code: Cursor, Claude Code, Codex, Devin, Aider. The category Skene is built to live alongside.

Coding agents

About this term

This page is part of the Skene analytics instrumentation glossary. Use it as a reference when writing specs, dashboards, or playbooks.

Canonical glossary index: /resources/glossary

Glossary

Glossary

Analytics callEvent namePayload (event properties)Identify callEvent taxonomyInstrumentation surfaceBaseline manifestInstrumentation driftSemantic diffRemoved eventRenamed eventMoved eventAltered payloadConditional firing changeCustomer Data Platform (CDP)Schema registryMCP (Model Context Protocol)Coding agentSide effectFunnelConversion eventActivation eventCohort

Definition

A coding agent is an AI assistant that operates on source code as its primary surface: reads files, writes diffs, runs commands. Some are conversational (Cursor, Claude Code). Some run autonomously over longer horizons (Devin). All of them are good at logic and indifferent to side effects.

Why analytics suffers under coding agents

Coding agents are pattern-matchers operating on logic. An analytics call is structurally invisible to logic: it does not return a value, it does not affect control flow, it does not change what the function computes. So when an agent rewrites a function, the analytics call inside is the part most likely to disappear.

Related terms

  • Side effect
  • Instrumentation drift
Skene

Product

How it worksFeaturesArchitectureIntegrationsSecurityPricing

Resources

DocumentationGlossaryPlaybooksBlog

Company

AboutOpen sourceContactPrivacyTerms
© 2026 Skene. All rights reserved.
Privacy PolicyTerms of Service
Skene