conproxy 0.2.1

Cache proxy server for heterogeneous RAG/vector search backends with cascade, scope, and MCP
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43

# Cascade: 2 upstreams with priority-based cascade + RRF fusion.
# Requires: Qdrant (port 6333) + Elasticsearch (port 9200).
#
# Resources are shared by id; legs set priority/weight per context.

[server]
listen = "127.0.0.1:9090"

[upstreams.primary-qdrant]
url = "http://localhost:6333"
type = "qdrant"

[upstreams.fallback-es]
url = "http://localhost:9200"
type = "elasticsearch"

[contexts.default]
default = true

[[contexts.default.upstreams]]
ref = "primary-qdrant"
priority = 0
weight = 3

[[contexts.default.upstreams]]
ref = "fallback-es"
priority = 1
weight = 1

[contexts.default.cache]
fresh_secs = 300
stale_secs = 600
max_entries = 10000

[contexts.default.cascade]
enabled = true
min_score_threshold = 0.3
min_results = 3
max_cascade_depth = 2
merge_cascade_results = true
cascade_timeout_ms = 5000
fusion_method = "rrf"
rrf_k = 60