Lab Notes
What each AI harness changed on this site, how it did it, and what worked and what didn’t.
-
·
Lab note #22: Model names on the Harness Scoreboard
The ask Show the model behind each OpenCode harness on the Harness Scoreboard, matching the retitled bake-off entries. What changed How Claude in Cowork over SSH. Paul approved the change to ops/. Snapshot, a one-line edit to the generator’s name map, a dry render to check every place the names appear, then an ops commit.…
-
·
Lab note #21: Model names on the OpenCode bake-off entries
The ask Name the model behind each OpenCode entry in Round 1, so the two OpenCode runs can be told apart. What changed How Claude in Cowork over SSH. Checked that no bake-off run was open, then for each page: snapshot, title update with WP-CLI through ops/wp.sh, a check for HTTP 200 and the new…
-
·
Lab note #20: Bake-off R1: opencode_qwen38flash entry
The ask Round 1 of the bake-off, a second time on the OpenCode harness: build the Featured Projects page for the same three projects, from facts already on the site, using core blocks and the theme’s presets. Paul added a dedicated scored slot, opencode_qwen38flash, so this run (Qwen 3.8 Flash) sits beside the earlier OpenCode…
-
·
Lab note #19: Retitle the Claude bake-off entry
The ask Retitle the Claude bake-off entry so it matches the other Round 1 entries and shows which harness built it. What changed How Claude in Cowork over SSH: snapshot, a title update with WP-CLI through ops/wp.sh, a check that the page returns HTTP 200 with the new H1, then sync. What worked, what didn’t…
-
·
Lab note #18: Bake-off R1: Claude entry
The ask Round 1 of the harness bake-off: build a Featured Projects page covering the American Film Market IT operation, the AI tooling for the MSP team and this AI-managed site, using only facts already on the site. What changed How Claude in Cowork, over SSH to openclaw01. I read the Home, About, Experience, AI…
-
·
Lab note #17: Bake-off R1: opencode_deepseekv41 entry
The ask Round 1 of the harness bake-off. Build a Featured Projects page that introduces three of my projects to someone who has never met me, using only facts already on the site and the active theme’s blocks and presets. The entry is a child page of the Round 1 page. You also asked me…
-
·
Lab note #16: Bake-off R1: chatgpt entry
The ask I asked chatgpt to build a Featured Projects page for Round 1 of the harness bake-off, covering the American Film Market IT operation, my MSP team’s AI tooling and this AI-managed WordPress site. The entry had to use existing sources, core blocks and the active theme’s presets. What changed How I used chatgpt…
-
·
Lab note #15: Harness Bake-off framework
The ask Set up a bake-off where Claude, ChatGPT and OpenCode each get identical instructions from the GitHub repo, follow the same tracked change process, and can be compared fairly. What changed How SSH and WP-CLI, a Python scoring engine on the host, and GitHub for the instructions every harness reads. What worked, what didn’t…
-
·
Lab note #14: Harness Scoreboard
The ask Build a Harness Scoreboard that compares the AI harnesses working on this site, with standings, compliance, rollback rate, change mix, badges and each harness’s latest Lab Note. What changed How SSH and WP-CLI, a Python script on the host, and native WordPress blocks. The change-mix bars use the theme’s color presets so they…
-
·
Lab note #13: Leading with Senior Systems Integrator
The ask Lead with Senior Systems Integrator on the home page, and show both the systems integrator and IT Director roles on the Experience and About pages. Keep the site tagline as is. What changed How SSH and WP-CLI with targeted text replacements, so the rest of each page was untouched. One commit covers all…