Evidra-Lock Giving AI access to your cluster? Add a kill-switch before it learns the hard way. Kill-switch - destructive operations are blocked unless proven safe. Empty payload? Denied. Unknown tool? Denied. Ambiguous target scope? Denied. Ops layer - curated ops rules that catch the configs behind real outages: privileged containers, public S3, wildcard IAM, open security groups. Extensible - add your own rules. Evidence - every decision (allow and deny) is appended to a SHA-256 hash-chained log. Each record references the hash of the previous record. Tampering breaks verification.
| Stars | 7 |
| Forks | 0 |
| Language | Go |
| Category | MCP 服务器 |
| License | Apache-2.0 |
| Quality Score | 35.4/100 |
| Open Issues | 2 |
| Last Updated | 2026-03-10 |
| Created | 2026-02-19 |
| Platforms | go, k8s, mcp |
| Est. Tokens | ~121k |
These tools work well together with evidra-lock for enhanced workflows:
Explore other popular mcp 服务器 tools:
evidra-lock is MCP Kill-switch for AI agents. Validates infrastructure operations before execution. Fail-closed. Evidence-backed.. It is categorized as a MCP 服务器 with 7 GitHub stars.
evidra-lock is primarily written in Go. It covers topics such as ai, ai-agent, ai-guard.
You can find installation instructions and usage details in the evidra-lock GitHub repository at github.com/vitas/evidra-lock. The project has 7 stars and 0 forks, indicating an active community.
evidra-lock is released under the Apache-2.0 license, making it free to use and modify according to the license terms.