api-relay-audit
@toby-bridges
Security
api-relay-audit is a local security audit tool for AI API relays and LLM proxies, detecting prompt injection, model substitution, tool-call rewriting, SSE anomalies, error leakage, and Web3 wallet risks.
All DeepSeek Harness plugins
19 pluginsGitHub data updated Aug 21, 2026
sandbase-harness
@sandbaseai
Security
sandbase-harness is a local-first AI agent runtime with sandboxed sessions, MCP tools, memory, credentials, audit/replay, and a built-in console, supporting OpenAI, Anthropic, MiniMax, DeepSeek V4, and OpenAI-compatible models.
dsh-pentest
@howmp
Security
Penetration testing mode for DeepSeek Harness (dsh) @CloverSecLabs.
dsh-auto-mode
@NanmiCoder
Security
Adds a middle ground between restricted and full access modes for DeepSeek Harness. Keeps Auto on the official workspace-write sandbox and adds review for semantic risks outside that boundary, asking once for genuine ambiguity and denying destructive access to critical paths.
dsh-undo-savepoint
@lire1131
Security
A crash-rescue plugin for DeepSeek Harness providing one-click rollback of config and plugin-code changes, secret-safe snapshots, one-click SAFE MODE, and offline CLI/GUI tools that work even when DSH won't boot.
dsh-turn-rewind
@Anionex
Security
Rewind conversation and workspace state in DeepSeek Harness, powered by a persistent Change Ledger.
dsh-auto-review
@PerryLink
Security
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.
dsh-claude-ux
@eri64
Security
DSH plugin: Claude-style Chinese risk control & conversation autonomy for DeepSeek Harness web
dsh-remote
@xgone
Security
Remote access for DeepSeek Harness: account/password auth + MFA (TOTP) login gate, signed session cookies, role-based access, in-browser directory picker, and a Settings page for account management.
dsh-plugin-guard
@lxzy-7
Security
Install safety net for DeepSeek Harness: pre-install snapshots, one-click/automatic rollback, guarded boot, and incident reports that auto-trigger agent analysis.
dsh-permission-rules
@PerryLink
Security
Claude Code-style declarative permission rules for DeepSeek Harness with ordered allow/deny/ask rules, tool-name and argument matching, session-log audit, and HMR reload.
dsh-full-remote
@JUANWANG-BUAA
Security
Auditable, token-gated DeepSeek Harness remote gateway with mobile QR access, per-device sessions, Host/Origin rewrite, and settings/credentials/directory support.