A pre-execution guard for AI coding agents. It blocks destructive Git and file system commands, plus common attempts to access sensitive files, before a tool call runs. Supports Amp Code, Antigravity CLI, Claude Code, Codex, Cursor, DeepSeek Harness, Devin CLI, GitHub Copilot CLI, Grok Build, Hermes Agent, Kimi Code, OpenClaw, OpenCode, and Pi.
Local security audit for AI API relays and LLM proxies: detects prompt injection, model substitution, tool-call rewriting, SSE anomalies, error leakage, and Web3 wallet risks. Website: https://apirelayaudit.com
SecurityDeveloperFinance
★ 911AGPL-3.0更新 10-27金
dsh plugin --profile web add dsh-api-relay-audit@2.5.0
Remote-work assistant for DeepSeek Harness (DSH): SSH into any machine, pick a remote workspace, and work on it with 22 rw_* tools, conflict-aware SFTP mirror, embedded PTY terminal, port forwarding, and multi-hop jumps - under the same sandbox policy as a local session.
Second-model AI auto-review for DeepSeek Harness approval requests: a read-only reviewer subagent returns structured allow/deny verdicts with reasons, fail-closed by default, fully auditable from the session log (approval/asked -> autoReview/verdict -> approval/decided).
DeveloperAI ModelsSecurity
★ 248Apache-2.0更新 11-09金
dsh plugin --profile web add dsh-auto-review@0.15.0