agentseal

5 ms·

1.
Eywa: Local-first memory for AI agents, with a receipt for every fact(arxiv.org)
2 points by agentseal 3mo ago|0 comments
2.
Where AI coding spend goes: 48% code, 40% thinking(codeburn.app)
2 points by agentseal 4mo ago|0 comments
3.
by agentseal 5mo ago
On the fewer tools per turn, yeah I think that lines up with what the other reply mentioned about 4.7 being more "in its head." I have not specifically tracked hallucinated project structure but the higher retry rate suggests it i
4.
by agentseal 5mo ago
The one-shot rate doesn't factor in context size directly, it just tracks whether an edit succeeded without retries. That said, a detailed CLAUDE.md probably helps both models equally since the context is the same either way. Would be
5.
Opus 4.7 vs. 4.6 after 3 days of real coding side by side from my actual session
16 points by agentseal 5mo ago|4 comments
6.
by agentseal 5mo ago
Noted!! I will create an issue on this and as soon as we provide a fix i will notify you here
7.
by agentseal 5mo ago
already on that probably in an hour it will be published. will let you know once it does.
8.
by agentseal 5mo ago
not $1,400 out of pocket, thats the API equivalent cost of the tokens. I am on the $200/month Max plan :D. In my case I mostly consume every bit of the weekly subscription.
9.
by agentseal 5mo ago
you are right. thats cursor-agent (the CLI), not the Cursor IDE. CodeBurn only parses the IDE's state.vscdb right now. cursor agent keeps transcripts under ~/.cursor/projects/*/agent-transcripts/ which we dont
10.
by agentseal 5mo ago
yeah Ink, it was the fastest path to something that felt native next to Claude Code itself
11.
Show HN: CodeBurn – Analyze Claude Code token usage by task(github.com)
112 points by agentseal 5mo ago|27 comments
12.
by agentseal 6mo ago
we scanned 50 popular shared .cursorrules files from github and found that 6 of them contained hidden zero-width unicode characters embedded between visible text these characters are invisible to humans but LLMs tokenize them individually,
13.
by agentseal 6mo ago
good question and your right that theres a real gap between "can the agent be tricked" and "can it actually cause damge once tricked." our current probes mostly test the first part. can we get the agent to leak credentia
14.
AgentSeal – Security scanner for AI agents (150 attack probes)(github.com)
1 points by agentseal 7mo ago|2 comments
15.
by agentseal 7mo ago
I built AgentSeal to answer a simple question: can your AI agent be hacked? It sends 150+ attack probes (prompt extraction, injection, persona hijacking, encoding tricks, etc.) at your agent and gives you a trust score from 0-100 with s