Codex builds a blog inside an existing Datasette site
Willison published the Codex session used to build the Datasette project blog. The agent inspected the existing site, moved conflicting database routes and implemented blog behavior with local checks.
gist.github.com
Suggested prompt
Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.
Add [content section] to this existing site. First trace URL conflicts and all callers of any route you must move. Implement a small slice using current conventions, verify local routes and preserve redirects where required. Present checks and a diff; stop before publishing.
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source author
- Simon Willison — published Codex session
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- The session documents implementation and local checks; Willison separately announced the new project blog.
- Human involvement
- Willison directed routing and product choices throughout the published session.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Inspect repository dependencies and verify a working local development server.
- Move the existing blog database route to free the desired path while updating callers.
- Implement the new blog in the existing site and run local verification.
More details
About this account
Codex traced route dependencies and edited related files, using commands and checks to guide implementation.
Publicly linked by the author from his May 13 blog. The transcript is deployment evidence; agent statements alone are not independent validation.
Credit: Simon Willison — published Codex session
Source published
Added to Agents In Use
Source checked
Tools and topics
Codex Desktop · Datasette · Python · uv · Git · curl
Code · Testing
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.