.files

bad I add it to the script. Only then do I update Claude Code. It was during one of these script maintenance sessions that I noticed the server side prompt injection mechanism. I'll also tell Claude to look for and disable this steganography nonsense from now on as well. I usually audit the environment variables too. > it seems like CLAUDE.md should have the final say I wouldn't count on it.

is. Here is the Wikipedia Article's first line (which I agree with, as a definition): > Prompt injection is a cybersecurity exploit and an attack vector in which innocuous-looking inputs (i.e. prompts) are designed to cause unintended behavior in machine learning models, particularly large language models (LLMs). Anthropic are sending down a system prompt to their proprietary software from their proprietary service. It isn't an exploit, isn't an attack vector, and isn't unintended or unexpected. > I can only tell you what I'm doing. Here you go: https://github.com/matheusmoreira/.files/blob/master/%7E/.lo... Those seem like pretty reasonable changes to the prompt. Why is altering the system prompt more effective than instructions after?

ure if this meets the common definition of "prompt injection." They're literally injecting strings from the network into the system prompt. If it's not prompt injection, then I have no idea what it is. > My gut tells me there is something else going on, leading people to hack the Claude Code prompt/binary. And that the "something else" isn't supported by Anthropic. No idea what others are doing. I can only tell you what I'm doing. Here you go: https://github.com/matheusmoreira/.files/blob/master/%7E/.lo...

s days to run a full agent sweep. I designed it around the rate limiting. The agents do file system style journaling in order to resume cleanly. They commit all of their findings as they go into an orphan branch in the repository. Further review runs can build on it and avoid searching for known issues. The way it works in practice is I just run /scrutinize sweep and then go work on something else, or just go do my actual job, live my life, play video games, write an article for my blog or something. Come back five hours later to either resume the process or check the literally hundreds of issues that have been found by all the agents. Then Claude and myself will go in and evaluate and fix all of those issues one by one. Then review again. Then evaluate/fix again. I'm just gonna keep looping this over and over until zero issues are found. For all of my projects. Going from solo hobbyist programmer to this was pretty insane. I can only imagine what these corporations with infinite money must be doing.

bst $(directories.source)/%.c,$(directories.build.objects)/%.o,$(1)) These functions are then augmented by lots and lots of project specific variables to organize things... directories.source := source directories.build.root := build directories.build := $(directories.build.root)/$(config) directories.build.objects := $(directories.build)/objects sources.program := $(shell find $(directories.source) -type f) objects.program := $(call source_to_object,$(sources.program)) Then I add a real target which is supposed to get the makefile going. targets.program := $(directories.build)/program $(targets.program): $(objects.program) | directories $(strip $(CC) $(CFLAGS) $(LDFLAGS) -o $@ $^) Then I add a phony target for it. targets.phony += program program: $(targets.program) .PHONY: $(targets.phony) Then typing "make program" triggers the rule to make build/$(config)/program which depends on all the object files which will be linked into it which in turn is derived from a dynamically computed list of source files, and making those object files triggers the pattern rule which builds the entire thing. Using make without phony targets is insane. Without phony targets, I'd need to type things like "make build/aarch64/program" in order to get a build started. So I use phony targets for everything. Much easier to type "make program". It got to the point I created a phony-targets shell script which parses make's database output and processes it into a sort of help text for any given makefile's phony targets interface: https://github.com/matheusmoreira/.files/blob/master/~/.loca...