The Claude Code Source Leak: fake tools, frustration regexes, undercover mode

Post by tacone » Thu, Apr 02, 2026, 10:20 AM UTC

tacone wrote: post #434

otterley wrote: post #41

Earlier quoted context omitted.

I would have expected people (maybe a small minority, but that includes myself) to have already instructed Claude to do this. It’s a trivial instruction to add to your CLAUDE.md file.

It doesn't work so well in my experience. I am currently wrapping (or asking the LLM to wrap) the commit message prompt in a script call. 1. the LLM is instructed on how to write a commit message and never include co-authorship 2. the LLM is asked to produce a commit message 3. the LLM output is parsed by a script which removes co-authorship if the LLM chooses to include it nevertheless

Post by drzaiusx11 » Thu, Apr 02, 2026, 11:46 AM UTC

graemep wrote: post #601

drzaiusx11 wrote: post #550

Earlier quoted context omitted.

Currently, the US copyright application process has an AI disclosure requirement for the determination of applicability of submitted works for protections under US copyright law. The copyright office still holds that human authorship is a core tenet of copyrightability, however, whether or not a submission meets the "de minimis" amount of AI-generated material to uphold a copyright claim is still being decided and re…

You do not apply for copyright. In the US you can, optionally, register a copyright. You do not have to, but it can increase how much you get if you go to court. I do not know whether any other country even has copyright registration. Your main point that this is something the courts (or new legislation) will decide is, of course, correct. I am inclined to think this is only a problem for people who are vibe coding.…

Post by d4rkp4ttern » Thu, Apr 02, 2026, 1:17 PM UTC

manwe150 wrote: post #598

d4rkp4ttern wrote: post #559

For me one of the most interesting aspects is how compaction works. It turns out compaction still preserves the full original pre-compaction conversation in the session jsonl file, and those are marked as "not to be sent to the API". Which means, even after compaction, if you think something was lost, you can tell CC to "look in the session log files to find details about what we did with XYZ". I knew this before the…

What is microcompaction? I didn’t realize there was any thing time based in CC, when I go eat dinner and come back it compacted while I was gone?

  The logic is:

  1. Anthropic's API has a server-side prompt cache with a 1-hour TTL
  2. When you're actively using a session, each API call reuses the cached prefix — you only pay
  for new tokens
  3. After 1 hour idle, that cache is guaranteed expired
  4. Your next message will re-send and re-process the entire conversation from scratch — every
  token, full price
  5. So if you have 150K tokens of old Grep/Read/Bash outputs sitting in the conversation, you're
  paying to re-ingest all of that even though it's stale context the model probably doesn't need

  The microcompact says: "since we're paying full price anyway, let's shrink the bill by clearing
  the bulky stuff."

  What's preserved vs lost:
  - The tool_use blocks (what tool was called, with what arguments) — kept
  - The tool_result content (the actual output) — replaced with [Old tool result content cleared]
  - The most recent 5 tool results — kept

  So Claude can still see "I ran Grep for foo in src/" but not the 500-line grep output from 2
  hours ago.

  Does it affect quality? Yes, somewhat — but the tradeoff is that without it, you're paying
  potentially tens of thousands of tokens to re-ingest stale tool outputs that the model already
  acted on. And remember, if the conversation is long enough, full compaction would have summarized
   those messages anyway.

  And critically: this is disabled by default (enabled: false in timeBasedMCConfig.ts:31). It's
  behind a GrowthBook feature flag that Anthropic controls server-side. So unless they've flipped
  it on for your account, it's not happening to you.

Post by flykespice » Thu, Apr 02, 2026, 2:59 PM UTC

pc86 wrote: post #585

andoando wrote: post #169

Earlier quoted context omitted.

I guess Im sometimes dishonest when it suits me

It's only dishonest not to include Claude in commit/PR attribution if it's also dishonest not to include StackOverflow, or VSCode, or VIM, or Windows, or any of the other tools you used to complete the work!

Post by riedel » Thu, Apr 02, 2026, 6:15 PM UTC

mynameisvlad wrote: post #607

riedel wrote: post #580

Earlier quoted context omitted.

Good question. Actually, i was assuming that at least source code is treated as text under the legal regime (there is typically special rules in copyright law, but provision applying to text should apply). Furthermore I would think pull requests, etc are all text. So I would think this applies.

But it's not just text. Once again, it's explicitly defined as: > which is published with the purpose of informing the public on matters of public interest There is no "informing public on matters of public interest" in source code nor an MR. It's clearly meant to prevent "deepfake" news, like the image and video ones explicitly call that out.