Infrastructure, AI and tech solutions.
We focus on solving problems and delivering finished products.
The products.
-
01
chat-recallchatrecall.dev
One searchable history for every AI coding tool on the machine.
- Temporal knowledge graph
- MCP server
- Secret alerts
-
02
gpuscalegpuscale.cc
Buys the cheapest spot GPU it can find, then gives it back.
- Seven providers
- Scales to zero
- vLLM
-
03
axonAsk us
One gateway for every LLM call, with the real cost attached to each one.
- Routes by cost and request type
- Response cache
- Per-tenant budgets
- Spend alerts
- Cost dashboards
-
04
codeindexgithub.com/munhq/codeindex
An agent calls a tool instead of reading the whole file.
- 40+ languages
- Dependency graph
- MCP server
-
05
cloud-toolsgithub.com/munhq/cloud-tools
Reads your cloud bill and says where the money goes.
- Four clouds
- Read-only access
- MCP + HTTP
-
06
distilgithub.com/munhq/distil
Measures what compressing a session actually costs.
- Benchmark harness
- Prompt-cache accounting
- MCP server
-
07
CoolCodecoolcode.hotmun.com
Runs coding agents as a team, with work that outlives a session.
- Desktop app
- Role-pinned models
- Every change snapshotted
-
08
munbotAsk us
A team of agents that runs a business, one isolated pod each.
- One pod per tenant
- Per-tenant spend ceilings
- Scheduled agent turns
-
09
agentfetchAsk us
Gets clean markdown out of pages that resist it.
- Three-layer escalation
- Markdown output
Write to us.
- General
- hello@munhq.com
- Support
- support@munhq.com