11 min read
I trusted AI agents after I stopped trusting their first draft
How deep tests, stacked pull requests, adversarial review and isolated environments let me delegate parallel software delivery safely.
Blog posts tagged
AIPage 1
7 min read
I'm fed up of people using the excuse of AI slop
AI slop is the result of weak direction and absent editorial judgement, not an excuse to blame the tool.
7 min read
How to Stop Reviewing Agent-Written Code
A multi-agent quality pipeline that replaces routine line-by-line review with executable evidence and deterministic gates.
9 min read
Why LLMs ignore the middle of your context
Long context windows do not guarantee reliable recall. Design retrieval, reranking and prompt assembly so the model can use the evidence it receives.
7 min read
Let an Open-Weight Model Audit Your Code, but Do Not Give It the Keys
How to run a nightly AI security audit without giving the model credentials, production access or authority to merge its own fixes.
7 min read
Claude's Watermarking Policy Extends to Code
Claude's invisible marks could improve provenance, but Anthropic has not yet shown that they leave generated code quality and optimisation untouched.
5 min read
Configure AI Models Instead of Hard-Coding Them
Keep model names in configuration, select models by task, and compare routes using the cost of completed work.
5 min read
The Pull Request Can Now Rewrite Its Own AI Reviewer
Why pull-request-controlled AI review instructions collapse a trust boundary, and how teams can restore it without giving up useful repository context.