lynkr

Version:

Self-hosted LLM gateway and tier-routing proxy for Claude Code, Cursor, and Codex. Routes across Ollama, AWS Bedrock, OpenRouter, Databricks, Azure OpenAI, llama.cpp, and LM Studio with prompt caching, MCP tools, and 60-80% cost savings.

fast-editor.github.io/Lynkr/

Fast-Editor/Lynkr

33 folders, 2 files

Filename	Content Type	Size
../
agents/	–	63.4 kB
api/	–	161 kB
budget/	–	16 kB
cache/	–	37.8 kB
clients/	–	221 kB
config/	–	52 kB
context/	–	76.1 kB
dashboard/	–	9.08 kB
db/	–	11.8 kB
diff/	–	3.53 kB
edits/	–	3.88 kB
headroom/	–	32.4 kB
indexer/	–	68.4 kB
logger/	–	47.6 kB
mcp/	–	21.6 kB
memory/	–	70.8 kB
metrics/	–	585 B
observability/	–	21.4 kB
orchestrator/	–	157 kB
policy/	–	20.9 kB
prompts/	–	13 kB
routing/	–	201 kB
server/	–	5.63 kB
sessions/	–	9.35 kB
stores/	–	3.54 kB
tasks/	–	8.34 kB
tests/	–	15.1 kB
tools/	–	168 kB
training/	–	8.38 kB
usage/	–	6.11 kB
utils/	–	20.8 kB
workers/	–	21.9 kB
workspace/	–	3.94 kB
cluster.js	text/javascript	2.77 kB
server.js	text/javascript	11.2 kB