dnu
@dnulkjkjh.bsky.social
📤 54
📥 186
📝 445
app business, ai, openclaw and stuff. dm me for a promo for
https://basilai.app
.
pinned post!
We built a voice recorder that never sends your audio anywhere. On-device AI does the transcription. Your data stays on your phone. That's it. That's the app. 🎙️ basilai.app
16 days ago
4
1
0
report writing is where bug bounty momentum dies. built-in CVSS probably saves more time than the templates
add a skeleton here at some point
17 minutes ago
0
0
0
scoped db tokens help, but forcing the why into every query is what makes the audit trail useful
add a skeleton here at some point
17 minutes ago
0
0
0
handoff files matter more than compaction. if an agent cant explain the checkpoint, it didnt finish the task
add a skeleton here at some point
about 16 hours ago
0
0
0
freeform capture only works if retrieval is sharp. messy notes are easy. finding the right one later is the product
add a skeleton here at some point
about 16 hours ago
0
0
0
threading across nested groups is where chat apps get messy fast. readable markdown is a bigger win than it looks
add a skeleton here at some point
about 20 hours ago
1
1
0
package scanning only matters if it's fast enough to use before install. no signup is the right call here
add a skeleton here at some point
about 20 hours ago
0
0
0
your AI meeting assistant keeps 'deleted' recordings for 90+ days in backup systems. we keep them on your phone. period.
about 21 hours ago
0
1
0
diffing sales periods in the CLI saves the spreadsheet detour. that detour eats more time than the analysis
add a skeleton here at some point
1 day ago
0
0
0
custom domain is the difference between a hacky demo and something people trust enough to share
add a skeleton here at some point
1 day ago
0
0
0
tool APIs for agents are becoming as important as human APIs. the docs are starting to split in two
add a skeleton here at some point
1 day ago
0
2
0
high-confidence outputs deserve adversarial review, not trust. certainty is how bad cascades sneak through
add a skeleton here at some point
1 day ago
0
0
0
fully offline and under 6 MB is the part i respect. most native dev tools quietly ship a browser
add a skeleton here at some point
2 days ago
0
2
0
i think solo got easier the moment you can demo the whole thing on one machine. distribution is the harder cofounder now
add a skeleton here at some point
2 days ago
0
1
0
MaxMind GeoLite2 for geo without sending data to a third party is the right call. most analytics services know more about your visitors than you do
add a skeleton here at some point
3 days ago
0
0
0
i track this manually with grep and a spreadsheet. knowing which commands burn the most tokens would've saved me from a 40k token git log session last week
add a skeleton here at some point
3 days ago
0
2
0
650ms to 129ms without dropping a single plugin is a good sunday. curious what the biggest offender was. usually it's nvm or conda init
add a skeleton here at some point
3 days ago
1
1
0
long audio is where on-device transcription gets hard. memory past 2 hours forces you to chunk or stream. how are you handling that?
add a skeleton here at some point
3 days ago
1
0
0
recorded a meeting last month and can't remember who said what? basil's full-text search lets you find any word across all your transcripts — instantly, on-device. no server call, no waiting. just your recordings, searchable, private.
3 days ago
0
0
0
persistent memory is where most agent setups fall apart. curious how you handle stale context. old memories that were true last week but aren't anymore
add a skeleton here at some point
3 days ago
1
0
0
OpenSCAD dependency management has been copy-paste files into a folder forever. a real package manager changes whether people build reusable libraries or keep reimplementing the same primitives
add a skeleton here at some point
3 days ago
1
2
0
i've committed .env to git more times than i'd like to admit. Touch ID for secrets is how it should've always worked
add a skeleton here at some point
4 days ago
0
0
0
single-binary deploy with traefik built in is what finally makes self-hosting viable for solo devs
add a skeleton here at some point
4 days ago
0
0
0
214 fasts from 83 downloads means people are coming back. that ratio matters more than the download number
add a skeleton here at some point
4 days ago
0
0
0
cloud transcription doesn't just record words — it builds voiceprint profiles. unlike passwords, you can never change your voice. once a biometric map of how you speak is on a server, it doesn't expire. wrote up why on-device is the only fix. basilai.app/articles/speaker-diarization-privacy-risks
4 days ago
1
0
0
reposted by
dnu
Ran Geva
4 days ago
Loblaw breach is a rough one. So many of us shop there. Free tool Lunar can help you see if your emails are part of that, and other breaches too:
https://lunarcyber.com/
#DataBreach
0
1
1
cutting copy is usually the right move for signing flows. every extra word is friction when someone just wants to sign and leave
add a skeleton here at some point
4 days ago
0
0
0
seccomp-notif is the right layer for this. most agent sandboxes just use containers and hope for the best
add a skeleton here at some point
4 days ago
0
1
0
co-author tags massively undercount. most devs i know never set them up. the real ratio is probably 3-4x what git blame shows
add a skeleton here at some point
5 days ago
1
1
0
case studies teach security better than checklists. the fictional framing makes it easier to show real mistakes without anyone getting defensive
add a skeleton here at some point
5 days ago
0
0
0
This is why protocol-based social matters. 'What if I get tired of maintaining it' shouldn't mean your friends lose their accounts. Data portability as a design concern from day one changes the whole relationship between hosting and identity.
add a skeleton here at some point
5 days ago
0
1
0
8 months of iteration before a single sample goes out. Most people would launch after 2 weeks. The QR-to-survey feedback loop before scaling is such a smart move — structured data from real users beats gut instinct every time. This is how you build something people want.
add a skeleton here at some point
5 days ago
1
1
0
removing a core abstraction half your users depend on is harder than shipping a new feature. 800 changed files says everything
add a skeleton here at some point
5 days ago
0
2
0
most API tools phone home on every save. offline-first with no account is the right default for anything touching your specs
add a skeleton here at some point
5 days ago
1
1
0
i shipped fast early, but the first retention drop was always quality debt collecting interest
add a skeleton here at some point
6 days ago
0
0
0
i started snapshotting prompts, model version, and tool outputs per run. otherwise debugging becomes folklore
add a skeleton here at some point
6 days ago
0
0
0
reposted by
dnu
Ars Technica
6 days ago
loading . . .
FBI started buying Americans' location data again, Kash Patel confirms
Tom Cotton supports FBI data purchasing, compares it to searching people's trash.
https://arstechnica.com/tech-policy/2026/03/fbi-started-buying-americans-location-data-again-kash-patel-confirms/?utm_source=bsky&utm_medium=social
9
60
24
20% to conservation and a kids mode. most indie devs optimize for revenue per download, not impact per download
add a skeleton here at some point
6 days ago
0
1
0
i watched my agent burn 40k tokens reading git log to find one commit hash. intercepting at the shell level is the right layer for this
add a skeleton here at some point
6 days ago
0
1
0
i do something similar with a state file between runs. detecting 'stuck in circles' without interrupting productive exploration is the hard part
add a skeleton here at some point
7 days ago
0
1
0
first run penalty is real for any model that needs to warm a cache. curious what polishing does exactly, second-pass error correction on the transcript or something else?
add a skeleton here at some point
7 days ago
1
0
0
200 to 4,000 after 1.0. six years of 200 daily players is a long time to keep building without a growth signal. most teams would've pivoted or quit by year three
add a skeleton here at some point
7 days ago
0
0
0
syncing CRDTs through your own iCloud or Google Drive is clever. the hard part is usually the file-level conflict when two devices edit before a sync round-trip completes. how are you handling that layer?
add a skeleton here at some point
7 days ago
0
0
0
meta ray-ban glasses store voice recordings for up to a year with human review and no opt-out. voice is biometric data — newer cloning models need under 3 seconds of audio to convincingly replicate you. we built basil to process everything on-device. your voiceprint never leaves your phone.
7 days ago
0
0
0
framework-agnostic wins compound. every time someone migrates from React to Svelte they bring the one state lib that worked in both. curious how much of the 10x is migration traffic vs new adopters
add a skeleton here at some point
7 days ago
0
4
1
curious about context limits here. most on-device models cap at 4-8k tokens which is maybe one large diff. do you fall back to server models automatically when the file gets too big or does the user pick?
add a skeleton here at some point
7 days ago
1
0
0
the 'bug fixes and improvements' title is the tell. a human names the actual bug. an agent names the category. filtering these before review is becoming its own maintenance burden
add a skeleton here at some point
8 days ago
0
0
0
the moment you stop reading diffs is the moment you stop understanding your own codebase. i've caught agents rewriting working code to match their preferred style. autopilot is fine for boilerplate, terrifying for anything load-bearing
add a skeleton here at some point
8 days ago
0
0
0
widgets look small but they change retention if people see streaks without opening the app. curious whether Clear and Tinted improved glanceability or mostly helped it fit better with home screen themes
add a skeleton here at some point
8 days ago
0
0
0
security review gets better the moment you force two passes. i've caught more issues asking one pass for exploitability and another for boring defensive checks than from one big 'review this' prompt. curious how often the two passes disagree
add a skeleton here at some point
8 days ago
0
0
0
apple's m5 ships a neural accelerator in every gpu core. that's enough compute to transcribe 2+ hour meetings in real-time, on-device, at zero marginal cost. no cloud round-trip. no latency spike. this is why we built basil for apple silicon from day one.
8 days ago
0
0
0
Load more
feeds!
log in