> Claude Code hooks are user-defined shell commands that execute at various points in Claude Code's lifecycle. Hooks provide deterministic control over Claude Code's behavior, ensuring certain actions always happen rather than relying on the LLM to choose to run them.
Claude Code released hooks a few days ago, and I set up a few hooks to improve my experience (and velocity) using Claude Code:
• check-package-age.sh: prevents CC from installing outdated packages
• code-quality-primer.sh / code-quality-validator.sh: Validates code quality
• code-similarity-check.sh: Helps prevent duplicate code (and uses an indexer to support faster method lookup)
• pre-commit-check.sh: Blocks bad commits with lint/test checks
• claude-context-updater.sh: Auto-updates CLAUDE.md to keep context fresh (and helps CC navigate the codebase faster)
To help me debug that the hooks were running, I added (local) logging, and I can see stats on the hooks that have run.
I tend to add a new hook when I notice CC not behaving as I want. I've got a backlog of hooks to add.
This is mostly functional, though before it being 'perfect' I figured I'd put it out here to get input and maybe some other contributors.
I have an O-1 with my current company where I’m a software engineer and not a founder. When I’m ready to leave to start my next business, what are my options to transfer it over? What would I need in place?
Hi Peter. Thanks for your prior help in our email dialog years ago about an e2.
If I’ve received my O-1 but I haven’t started working at the company yet, can I still travel to the US on my O-1 while working remotely for a canadian company?
They have done causal research and they have hidden it from FBers asking for results of the study.
Members from the well-being team tried to conduct research but was refused do to it because of “legal and PR” reasons. The well-being team was shut down.
All this will come out in lawsuits if/when they happen. Likely through attorney generals and not individuals.
Location: San Francisco | Remote | No relocation
Technologies: node, react, vue, swift, python, mongo and open to learning more.
Résumé/CV: https://drive.google.com/file/d/1BcGPFWgpamRXXBXXxMSXHxlrvKE6eH2I
WHY TO HIRE DAN:
Former founder, both iOS apps I've worked on were featured on the front page of the App Store globally, used by 300k+ people
3 yrs as a full stack product engineer and 2 yr managing engineers
Strong desire to level up as an engineer, and very receptive to feedback
My highest personal value is awareness, which extrapolates into every part of my personal and professional life - I'm a mindfulness nerd.
WHY NOT TO HIRE DAN:
You're looking for someone with strong systems design and architecture experience. Most of my experience is working within a pre-existing architecture (front and back end) and building front-end features and their corresponding end-points
I don't have experience working with an engineering team larger than 4 people
You don't have capacity to have a pair programming session once a week with me to help me level-up as an engineer
You're looking for a senior engineer
You expect the engineer to be highly proficient with SQL or GraphQL vs. learning on the job
You haven't raised a series-a yet (or on a strong trajectory to do so)
Or simply that one doesn't want to work for a company powered by hopes & dreams while praying for more cash. That early stage stock grant doesn't pay that SF rent.
Even with a raised round, the units that you are denominating the calculation is months. Months of life before your company dies (you can’t raise debt if you can’t raise equity, simplistically).
Depending on your risk appetite, family factors and time in life, it seems perfectly reasonable to signal a better value alignment. He/she could take a (let’s say gigantic) equity stake in a pre-seed entity and low cash, in which case it’s be an extremely expensive employee with no sensitivity around the round or stage of company.
Also, and sorry for belaboring this, but having a round starts the “clock”; personally that pressure to do X or raise Y or sell/IPO is not a feature but a bug, and for others that common goal orientation and team alignment might be really valued.
Seed funding: There is little to no revenue. There is no business model. The startup is trying lots of different things to see what sticks. Most of the work is building prototypes and throwing them away.
Series A: Startup is making money and has a repeatable and scalable business model. They may still be in the red with fixed costs, but every dollar spent on marketing returns more than $1 in income. The work is usually turning those hacky prototypes into stable code.
Location: SF
Remote: yes
Relocate: unlikely
Technologies: node, react, vue, swift, and open to learning more.
Resume: https://www.dropbox.com/s/v4oirwk9cy8ahfz/Dan%20Seider%20Res...
Email: in resume
Why me?
Former technical founder so I’m used taking a lot of ownership. A strong product sense: both apps I led and worked on were featured by Apple globally. Hungry to learn.
Why not me?
I’m not a senior engineer and I’m looking for a place that can offer some mentorship and the occasional pair programming session, so if you can’t offer that, we’re unlikely a fit.
Ad blockers are used by 600m devices.
40% desktop
15%. Mobile
(at least in the US)
Happy to gather sources if you'd like, I searched this up a few days ago
Claude Code released hooks a few days ago, and I set up a few hooks to improve my experience (and velocity) using Claude Code:
• check-package-age.sh: prevents CC from installing outdated packages • code-quality-primer.sh / code-quality-validator.sh: Validates code quality • code-similarity-check.sh: Helps prevent duplicate code (and uses an indexer to support faster method lookup) • pre-commit-check.sh: Blocks bad commits with lint/test checks • claude-context-updater.sh: Auto-updates CLAUDE.md to keep context fresh (and helps CC navigate the codebase faster)
To help me debug that the hooks were running, I added (local) logging, and I can see stats on the hooks that have run.
I tend to add a new hook when I notice CC not behaving as I want. I've got a backlog of hooks to add.
This is mostly functional, though before it being 'perfect' I figured I'd put it out here to get input and maybe some other contributors.