Dev Log

These entries change as providers publish new evidence. Their updated date tells you when we last checked the claim—not when the story first appeared.

• watching

  Is Claude watermarking your code yet?

  Anthropic gave its content-marking rollout a precise cutoff: 2 August 2026. It is a model-launch cutoff, not proof that every current Claude response is already marked.

  Published 16 August 2026 Last updated 16 August 2026

  Claude provenance regulation

No watched concepts match that filter.

Fixed records of a specific investigation, published when the work is done.

• The canary that catches what --version can't

  Codex 0.144.0 installed cleanly, printed the right version, and could not execute a single shell command. What the release broke, and the three layers built so the next one is caught first.

  Published 10 July 2026 Last updated 16 August 2026

• What a synthetic session store does and doesn't reproduce

  Benchmarking startup against a real home measures one machine's history. A synthetic store fixes that — and reproduces Claude Code's response latency within 10% while missing Codex's by 39%.

  Published 30 June 2026 Last updated 16 August 2026

• Three numbers for one context window

  Claude Code assumes 200K tokens for any model it doesn't recognise — including whatever your proxy is serving. Finding the right number for GPT-5.6 meant reading the binary, and the answer turned out not to be the one in the API docs.

  Published 10 August 2026 Last updated 16 August 2026