Repository Issue Activity (beta)

#136 Dependency Dashboard

#2910 Status of the project

#2908 [packit] Propose downstream failed for release v0.24.0

#2906 Add support for --name for RAG build, propagate names through the stack

#2847 change default transport?

#2837 serve --rag: streaming responses use content-type: text/plain instead of text/event-stream — strict OpenAI clients (e.g. Open WebUI) reject the stream

#2897 --generate` (quadlet/kube/compose) silently disabled in Toolbox and Distrobox — worth calling out in immutable-host install docs

#2880 ramalama daemon command docs incorrect

#2867 Missing FFmpeg dependency in the Llama.cpp container

#2836 rag: doc2rag counts chunk tokens with cl100k_base but the embedding server rejects >512 real tokens — "input (N tokens) is too large to process"

#2859 RFE: OCI artifacts for skills, agents and plugins

#2764 KleidiAI CPU backend isn't compiled by default for asahi hosts

#2798 ramalama does not pass through GPU when run in a toolbox(1), though llama.cpp does

#2831 `rag` erroring with "embedding dimension unknown"

#2878 perplexity has several small issues

#2819 "bad CPU type in executable" when running in Intel mac

#2875 README.md links are broken on github

#2823 `--backend rocm` is ignored for `ramalama bench`

#2860 Ramalama does not correctly say nvidia toolkit is not installed

#2813 Depricate Ramalama chat in favor of a default agent

#2816 Ramalama sandbox enable multi model support

#2806 Investigate Multi Modal Support for Ramalama

#2818 uv: command not found

#2843 ramalama convert failing on main

#2800 serve -d to host other than 127.0.0.1 or 0.0.0.0 causes health check timout