Skip to content
The Daily Repo

acodercat / cave-agent

Stateful runtime management for LLM agents—inject, manipulate, and retrieve Python objects across turns.

Why this job: topics: llm-agent, llm-agents

What it does

  • Code-based function calling. the LLM writes and runs Python against injected objects, not rigid JSON tool schemas
  • Stateful runtime. inject/retrieve real Python objects (DataFrames, connections, class instances) with state persisting across turns; expose class schemas for type-aware code generation
  • Two runtimes. in-process IPythonRuntime or crash-isolated IPyKernelRuntime
  • Reasoning models. streams o1 / DeepSeek-R1 reasoning live as ThinkingChunkEvents
  • Execution control. step, wall-clock and cumulative-token budgets, plus per-execution / stream-idle timeouts

Get started

pip install 'cave-agent[all]'
Maintenance
Active
Last commit 55 days ago
Stars
219
+14 this week
Licence
MIT
Allows business use
Downloads for
None listed
No release files
Latest version
v0.8.0
Released 3 months ago
Built with
Python
100% of the code

Is it looked after?

Contributors
2
Last 90 days
Issues
100% closed
0 still open
Pull requests
26 merged
0 waiting
Releases
25
10 in the last year

Similar repos

Repos similar to this one
RepoStarsLast commitLicenceDownloads for
The-Pocket / PocketFlow-Tutorial-Codebase-Knowledge12,7455 months agoMITNo downloads
neuron-core / neuron-ai2,1694 days agoMITNo downloads
monocle2ai / monocle3453 days agoApache-2.0No downloads
More: languages, topics and project record

Languages

  • Python 100%

Topics

  • llm
  • llm-agent
  • llm-agents
  • llm-evaluation
  • llm-inference
  • llm-tools
  • llms

Project record

  • JobAgent frameworks
  • StatusListed
  • Commits118
  • Forks18
  • Watching3
  • Stars in 28 days+15
  • CreatedJan 23, 2025
  • First seen by usNov 7, 2026
  • CheckedNov 9, 2026

Type to search.

Try