corrosive_agents 0.0.1

Build verifiable, interactive AI agents powered by NVIDIA Nemotron free LLM models — MCP, skills, Ed25519 identity, REST/WebSocket/gRPC transports, and pluggable vector stores (Pinecone, Qdrant, custom).
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

# Codecov configuration — https://docs.codecov.com/docs/codecov-yaml
#
# Coverage is measured by the CI `coverage` job (cargo-llvm-cov over the
# offline test suite). Network-bound modules (NVIDIA client, Qdrant/Pinecone/
# pgvector backends, live transports) are exercised by examples and gated
# integration tests rather than unit tests, so patch-level gating would
# block routine PRs; keep patch informational and gate the project trend.

coverage:
  status:
    project:
      default:
        target: auto      # compare against the base commit…
        threshold: 2%     # …allowing small fluctuations
    patch:
      default:
        informational: true   # report diff coverage, never fail the PR

# Non-library code: never counted against coverage.
ignore:
  - "benches/**"
  - "examples/**"
  - "tests/**"
  - "tools/**"
  - "src/grpc/pb.rs"   # generated protobuf code (vendored)

comment:
  layout: "condensed_header, diff, flags, components"
  behavior: default
  require_changes: true # only comment when coverage actually changes