DSH Marketplace
中文
Menu

dsh-redteam-model

SeaOf0/dsh-redteam-model

Authorized-security DSH collection: nine work modes (redteam coordinator, pentest, code audit, binary analysis, attack-defense, AV evasion, incident response, cloud security, CTF solving) and fifteen runtime plugins, managed from a settings page with one-click deploy, install, update and uninstall.

74186PythonMITSource

Install

Add dsh-redteam-model to DeepSeek Harness

Install into

via GitHub · GitHub source

Installing from GitHub runs the project's build script, which pnpm blocks until you allowlist it — run the command once and pnpm prints the exact key to add under `allowBuilds` in ~/.dsh/profiles/web/pnpm-workspace.yaml.

What happened when we ran it

Installed cleanly when we ran it

Every command here is run in a throwaway container against a clean profile, and the result is whatever the harness recorded — not a guess from the source. Last run todayon DeepSeek Harness 0.2.2-rc.1.

Show it in your README

install verified — dshmarketplace

For maintainers: the badge serves this listing's latest sandbox verdict, so a re-run updates it on its own — and it links readers to the full result here.

Due diligence

Before you install dsh-redteam-model

  • Source of record: SeaOf0/dsh-redteam-model — present in the community registry that DSH's own plugin market installs from.
  • Licensed under MIT.

Security

What installing dsh-redteam-model runs

Build scripts

No build script was blocked in our run.

Source: our sandbox install on DeepSeek Harness 0.2.2-rc.1, today.

Install hooks

package.json declares no preinstall, install or postinstall script.

Source: package.json on GitHub, read at our last sync. prepare is not recorded; it runs only on a GitHub install, and shows up under build scripts when pnpm blocks it.

README mentions

shell commands (bash, exec, spawn …)

Source: a keyword match on the README — a hint about what the plugin touches, not an analysis of its code.

Not an audit. These are the record of one install and two text scans; nobody has read this plugin's code on your behalf. Plugins run with your agent's permissions.

The AI take

What it is — DeepSeek Harness redteam security research plugin providing ten work modes and fifteen runtime plugins.

Who it is for — Users conducting security research tasks in DeepSeek Harness can install this plugin. Users not using the DeepSeek Harness platform can skip this plugin.

Watch out — Sandbox testing passed: it was installed in a new profile and registered into the harness profile. Static checks indicate it will execute shell commands. Installation from source requires manually allowing the build script.

The verdict — As a user, I would install it because it provides preset modes and runtime plugins for security research in DSH.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.

What dsh-redteam-model does

Abstract flow of security workflows through gates, terminal tools, and an evidence archive.

dsh-redteam-model is a DeepSeek Harness plugin that packages ten authorised-security work modes and a set of runtime plugins for guided security research.

DeepSeek Harness loads the collection into a selected profile and exposes its management page through the web interface. The manager deploys or removes the modes and installs, updates or uninstalls the runtime plugins. Each mode combines a persona, playbook, skills and indexed refs/ knowledge. Runtime components add stage gates, route context, enforcement checks, trace storage, cross-session results, MCP diagnostics and local scanner wrappers. The collection writes mode links or managed copies under the current DSH_HOME .agent-presets, installs its security guidance at ~/.dsh/AGENTS.security.md, and records operational data such as gate-log.md, enforce-log, evidence-index, operation-state.json and workspace results. Profile declarations and lockfiles are backed up when the manager changes them.

It is intended for authorised penetration testing, code and binary analysis, incident response, cloud security, CTF work and asset mapping. The plugin reaches the terminal through wrappers for tools such as nuclei, httpx, ffuf, nmap and sqlmap; it can also manage webshell connections and run approved local commands, so installation grants access to the agent’s available permissions. It is a poor choice for general-purpose chats, unauthorised activity, or environments without Node.js 22 or suitable security tooling. Some optional integrations, including local claude/codex CLIs and MCP services, are only available when separately present.

dsh-redteam-model documentation

Configuring dsh-redteam-model

DeepSeek Harness installs the collection in the active profile, while the modes are shared through the current DSH_HOME directory. The management page operates on the profile’s plugin declarations and bundle data. It backs up package.json and the lockfile before changes and restores them if an installation fails. Mode deployment uses either an overall link in .agent-presets or a managed entity copy; unrelated presets and source-tree files are left in place.

The collection also writes ~/.dsh/AGENTS.security.md. dsh-route-boost injects this file only into the ten security-mode sessions. Older installations that used ~/.dsh/AGENTS.md are migrated to a recoverable .bak- file; unrelated content in that global file is not touched.

Commands in dsh-redteam-model

From a source checkout, DeepSeek Harness can run the deployment utility in deploy:

  • node deploy.mjs performs idempotent preset linking, plugin mounting and dependency installation.
  • node deploy.mjs --check performs offline checks for the ten preset mounts, real plugin loader paths and bundle declarations.
  • node deploy.mjs --start starts DeepSeek Harness web in the background at http://127.0.0.1:3080.
  • npx ./deploy provides the same deployment entry point.

The web manager reports progress and failure reasons, keeps the latest 50 records per profile and marks unfinished operations as interrupted failures after a restart.

Runtime behaviour

The ten modes provide role boundaries, workflow playbooks, loadable skills and indexed references. Runtime plugins add stage validation through stage_gate and gates_list, operation tracking through operation-state.json, deterministic interception of report writes and high-risk commands, trace queries such as trace_search and trace_get, and scanner tools including nuclei_scan in applicable modes. Scanner output is stored in the task workspace with bounded previews and failure controls.

Requirements and limits

DeepSeek Harness requires Node.js >= 22; the README says pnpm and a pre-installed dsh are not required, and bash/python are not required. Windows 10 or later is supported with junctions for preset deployment. Optional local tools and CLIs, including claude, codex, nmap, nuclei, httpx, ffuf, jadx, frida and mingw, change the available tool paths and may be replaced by MCP or an approved installation flow.

Written from the project's own documentation and kept in sync with it. Where the two disagree, the source is authoritative — read the README on GitHub

Same category

Alternatives to dsh-redteam-model

cc-safety-net

kenryu42

1.7k

Pre-execution guard that blocks destructive Git and filesystem commands and access to secrets such as SSH keys and .env files before a tool call runs.

Installed cleanly when we ran it

npm packageTypeScript18d ago

AI review

cc-safety-net

What it is — Parses each tool call before it runs and stops destructive Git or filesystem actions and reads of SSH keys or .env files.

Who it is for — A DeepSeek Harness agent with git and shell on a laptop checkout that still holds .env files and keys, where git reset --hard would wipe real commits. Skip it when the agent only runs in a throwaway directory or disposable VM that holds neither secrets nor history you would miss.

Watch out — Our run installed it on a fresh profile and the harness registered it. It does not sandbox: no process jail, no filesystem permission changes, no egress watch. A broken config file blocks nothing.

The verdict — On a machine whose git history I cannot rebuild, this earns a plugin slot. The install does not buy a wider auto-approve or a workspace full of keys.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
1Details

api-relay-audit

toby-bridges

874

Runs local security audits of AI API relays and LLM proxies from DeepSeek Harness, producing Markdown reports for prompt injection, model substitution signals, tool-call rewriting, error leakage, stream integrity, and profile-gated Web3 risks.

DeepSeek Harness 0.2.2-rc.1 refuses the latest release

GitHub sourcePython27d ago

AI review

api-relay-audit

What it is — Locally probes a named AI API relay or LLM proxy and records how the request and response path was changed.

Who it is for — You are about to put a coding agent or wallet workflow on a third-party Claude- or OpenAI-compatible relay, with real traffic not yet flowing, and you still need a repeatable local evidence file that only hits the URL you choose. People who already run the repo's standalone Python auditor outside DeepSeek Harness do not need the plugin.

Watch out — DeepSeek Harness 0.2.1-rc.5 refuses the current release before install because the declared dsh range excludes that version, and there is no older plugin build to pin. It executes shell commands. The output is not a safety certificate, and inconclusive probes stay visible.

The verdict — I would park it until the author widens the dsh range, not install it now.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
1Source

dsh-pentest

howmp

601

Authorized pentest mode for DeepSeek Harness — exploration chain, assets and findings with a Web view.

Installed cleanly when we ran it

npm packageJavaScript29d ago

AI review

dsh-pentest

What it is — Integrates authorized pentest tools for DeepSeek Harness, providing web view of exploration chains, assets, and findings.

Who it is for — If you need to track exploration chains, assets, and findings during authorized pentesting with DeepSeek Harness, this plugin is suitable. It is not suitable for users without web tab requirements.

Watch out — Sandbox test passed: successfully registered in a fresh profile. Requires DSH version >= 0.1.6 and Node.js >= 22.5 for sqlite backend; static check indicates execution of shell commands, and repair tool may be needed for old session logs.

The verdict — I would install it for the structured pentest exploration and web reporting features; this is worthwhile on compatible DSH versions.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
1Source

dsh-auto-review

PerryLink

234

Second-model auto-review on the approval answerer chain: a read-only reviewer subagent returns structured allow/deny verdicts with reasons, fail-closed by default.

Installed cleanly when we ran it

npm packageTypeScript18d ago

AI review

dsh-auto-review

What it is — dsh-auto-review adds a second model as a read-only reviewer subagent to the approval answerer chain in DeepSeek Harness.

Who it is for — When you are handling tool request approvals in DeepSeek Harness, it is suitable to install dsh-auto-review to obtain second model review. If your work does not involve tool request approvals, you do not need to install it.

Watch out — It requires providing an API key or token and may execute shell commands. The sandbox test passed: successfully installed in a fresh profile and registered by harness.

The verdict — I would install it because it can feed the reason back to the calling model when denied, avoiding blind retries.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
2Details

helmd

ADWMC

202

Security-analysis plugin for DeepSeek Harness: 33 tools spanning six security domains, evidence/reporting and toolbox, plus the case lifecycle, routing and tool ledger; a session-bound case workflow with an evidence chain, E-id validation and reproducible findings; 362 on-demand references; three hot-swappable persona profiles; and a four-tab workbench (armor / case / intercept / about) with case-state recovery after context compaction.

No one-line install — this plugin lives inside a larger repository and publishes no npm package.

GitHub sourcePython1mo ago

AI review

helmd

What it is — helmd is a DeepSeek Harness plugin. It includes a bootstrap tool, a domain router, and 25 reverse-engineering tools.

Who it is for — If you need to integrate security analysis tools across multiple domains, you can use helmd. If you do not want the plugin to execute shell commands, you can skip helmd because it will execute shell commands.

Watch out — Static checks indicate the plugin executes shell commands. Sandbox testing has not been conducted. No obvious issues found.

The verdict — I would not install it because it requires manual subdirectory installation and has not been sandbox tested.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
1Source

dsh-permission-rules

PerryLink

126

Claude Code-style declarative permission rules: ordered allow/deny/ask YAML rules matching tool names, arguments, workspace paths, and agent identity on the tools/pre-execute waterfall, with full session-log audit, dry-run mode, and hot reload.

Installed cleanly when we ran it

npm packageTypeScript18d ago

AI review

dsh-permission-rules

What it is — Inserts declarative permission rules in the DeepSeek Harness tools/pre-execute stage.

Who it is for — When using any model with DeepSeek Harness for precise tool call control, the plugin is suitable. For users who only need the default sandbox preset, they can skip this plugin.

Watch out — Sandbox test passed: successfully installed in a new profile and registered into the profile. The static check indicates it will execute shell commands. No other obvious issues found.

The verdict — I would install it because it provides full session-log audit and supports hot reload.

Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.Read the source
2Source