AgentGuard
Dev tools
api cost monitoring
SHOW HN2025
How HN took it
0 positive, 2 neutral or asking, 0 mixed, 15 critical, out of 17 comments about the project. Thread score -0.68.
Execution: 0 for, 10 againstIdea: 0 for, 3 againstDesign: 0 for, 2 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Idea of a budget circuit breaker for AI costs is good.
From 1 comment
So it monkey-patches a set of common http libraries and then detects calls to AI APIs? Not obvious which APIs it would detect or in what situations it would miss them. Seems kind…
almost on HN
Objections
Monkey-patching HTTP libraries is intrusive and fragile; unclear which APIs are detected.
From 2 comments
So it monkey-patches a set of common http libraries and then detects calls to AI APIs? Not obvious which APIs it would detect or in what situations it would miss them. Seems kind…
almost on HNSo this is essentially monkey-patching every variation of fetch/library fetch and doing math on the reported token counts? It's an... intrusive solution. Glad to hear it works…
yifanl on HN
A proxy server such as LiteLLM already does this better.
From 3 comments
Yeah I don’t understand this problem, who uses so many agents at the same time in the first place? And if this is really a problem, why not funnel your AI agents through a proxy…
stingraycharles on HNIn fact, proxy servers like LiteLLM[1] already exist that are very easy to setup and set limits in. [1] LiteLLM: https://www.litellm.ai/
sothatsit on HNFor this I use LiteLLM proxy - you can create virtual keys with daily/weekly/... budget, its pretty flexible and has a nice UI. See https://docs.litellm.ai/docs/proxy/users
samsk on HN
Code appears vibe-coded with overstated README claims.
From 3 comments
Honestly feels very vibe-coded [1] [2] and would not really trust my money with something like this. I had to read the code to understand what it actually protects me from, as…
jeffhuys on HN"Commit 2ef776f dipampaul17 committed Jul 31, 2025 · Update READMEs: honest, clear, aesthetic - Removed pretentious language and marketing speak - Added real developer experience…
hansmayer on HN> [1] https://github.com/dipampaul17/AgentGuard/commit/d49b361d7f345ddba538deebcda9e23a2cdaeaa8 It's kind of crazy that people use these multi-billion parameter machine learning…
diggan on HN
Requests
Make it clear how and when it fails to detect calls.
From 1 comment
So it monkey-patches a set of common http libraries and then detects calls to AI APIs? Not obvious which APIs it would detect or in what situations it would miss them. Seems kind…
almost on HN
Questions
Why not use agents that respect usage limits instead?
From 1 comment
Wouldn't the obvious solution to this problem to stop using agents that don't respect your usage limits instead of trying to build sketchy containers around misbehaving software?
throwaway_ocr on HN
Alternatives named
LiteLLM proxy with virtual keys and budgets
From 2 comments
For this I use LiteLLM proxy - you can create virtual keys with daily/weekly/... budget, its pretty flexible and has a nice UI. See https://docs.litellm.ai/docs/proxy/users
samsk on HNIn fact, proxy servers like LiteLLM[1] already exist that are very easy to setup and set limits in. [1] LiteLLM: https://www.litellm.ai/
sothatsit on HN
tokencost for measuring costs
From 1 comment