Evidra

io.github.vitas/evidra·v0.4.7·Other

Quality Score

/100

Evidra — Flight recorder for AI infrastructure agents

§01 Install

Claude Desktop (Docker)

{
  "mcpServers": {
    "evidra": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "ghcr.io/vitas/evidra-mcp:0.4.7"
      ],
      "env": {
        "EVIDRA_ENVIRONMENT": "<evidra_environment>",
        "EVIDRA_EVIDENCE_DIR": "<evidra_evidence_dir>",
        "EVIDRA_SIGNING_MODE": "<evidra_signing_mode>",
        "EVIDRA_SIGNING_KEY": "<your-evidra_signing_key>",
        "EVIDRA_SIGNING_KEY_PATH": "<evidra_signing_key_path>"
      }
    }
  }
}

§02 Environment variables

EVIDRA_ENVIRONMENT

Environment label (production, staging, development)

EVIDRA_EVIDENCE_DIR

Directory for evidence JSONL storage

EVIDRA_SIGNING_MODE

Signing mode: strict (default) or optional (ephemeral local mode)

EVIDRA_SIGNING_KEY

secret

Base64-encoded Ed25519 private key (required for strict mode)

EVIDRA_SIGNING_KEY_PATH

Path to PEM-encoded Ed25519 private key (required for strict mode if key not inline)