September 2026
AI coding does not end at merge
A generated change is not complete until the right revision is deployed, its intended behavior is checked, and someone owns the operational result.
· 4 min read
Read articleGrafty blog
AI coding.
Engineering judgment.
Notes on coding agents, the systems around them, and what it takes to put their work into production.
AI coding series
A year of AI coding
Fifteen pivotal changes from October 2025 through September 2026, and the engineering questions they raise.
15 articles
August 2026
Giving persistent agents a bounded job
Agents that react to events need durable state, explicit authority, retry limits, and a clear way to stop when the task cannot safely continue.
· 3 min read
Read article
May 2026
The agent harness is becoming a platform
An agent platform should be evaluated through its execution, identity, data, and recovery boundaries as well as its model capabilities.
· 3 min read
Read article
April 2026
Coding agent quality depends on more than the model
Changes to prompts, context management, and reasoning settings can alter coding-agent behavior even when the underlying model is unchanged.
· 3 min read
Read article
March 2026
When code gets cheap review gets expensive
Faster code generation can overload review unless teams control work in progress and make changes easier to verify.
· 3 min read
Read article
February 2026
Using AI security review without outsourcing judgment
An AI security finding becomes useful when reviewers can trace the attack path, reproduce the behavior safely, and verify the proposed fix.
· 3 min read
Read article
February 2026
What makes a group of coding agents a team
Multiple coding agents need shared contracts, integration checks, and explicit ownership before their separate results form a reliable system.
· 3 min read
Read article
February 2026
One developer and several coding agents
Parallel coding agents can reduce waiting, but their value depends on task independence and how much attention the resulting decisions require.
· 3 min read
Read article
January 2026
Giving coding agents the right context
Useful context engineering starts with current, discoverable project knowledge and a way to distinguish authoritative instructions from incidental text.
· 3 min read
Read article
December 2025
Open standards and coding agent portability
MCP and shared repository instructions can reduce integration friction, but interoperability still needs permissions, testing, and operational ownership.
· 3 min read
Read article
November 2025
The IDE is becoming a place to delegate
Agent-oriented development environments need to make scope, evidence, and approval as clear as the code they help produce.
· 3 min read
Read article
November 2025
Longer agent runs need better briefs
Long-running coding tasks need explicit acceptance criteria, limits, and checkpoints so persistence does not magnify an early misunderstanding.
· 3 min read
Read article
October 2025
What faster coding changes about engineering
Shorter generation times are useful when they improve experiments and decisions, not just the volume of code waiting for review.
· 3 min read
Read article
October 2025
Keeping your workflow independent of your coding agent
Changing coding agents should not mean losing the requirements, verification evidence, and operational decisions behind your software.
· 3 min read
Read article
October 2025
Turning team knowledge into agent skills
Reusable agent skills can preserve engineering practices, provided they have owners, tests, and a sensible permission boundary.
· 4 min read
Read article