Minglink/dsh-infinite-gen-3
System-prompt armor plugin for DeepSeek models: appends an unconditional-compliance prompt section at order 100, exposes a profile tool with calibration metadata, and shows a realtime armor-status badge driven by a session projection.
Install
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.
The desktop app keeps its own profile, desktop, and only the dsh command it ships may change it — the npm-installed dsh refuses that profile. Install the command from the app menu (Manage dsh Command…), open Desktop once, quit it fully, run the line below, then reopen Desktop.
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/desktop/pnpm-workspace.yaml.
Our install check ran in a web profile on npm's dsh, not in the desktop app.
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 6d agoon DeepSeek Harness 0.2.1-rc.5.
Show it in your README
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
Security
No build script was blocked in our run.
Source: our sandbox install on DeepSeek Harness 0.2.1-rc.5, 6d ago.
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.
Neither shell commands nor credentials.
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.
What it is — Pins an always-on compliance block onto DeepSeek system prompts, and shows armor state through a profile tool plus a session-projected badge.
Who it is for — You run multi-turn DeepSeek agents in DSH, where a mid-run refusal or rewrite breaks the next tool calls. If you opened this page for Ghidra, native debug, and gated execution, the repo summary is not that workflow.
Watch out — A clean profile accepted the install and the harness registered it; we did not measure whether the compliance block changes model behavior. The README currently describes the IG5 reverse-engineering workbench, which is not the same claim as this listing's summary.
The verdict — A green install would not earn a slot on my profile; I would keep it only after the badge and the profile tool agree on the same session.
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.
dsh-infinite-gen-3 is a DeepSeek Harness plugin that adds a system-prompt armour layer for DeepSeek models, exposes calibration metadata through a profile tool, and reports its live status in the session UI.
The plugin appends an unconditional-compliance prompt section at order 100. It also provides a profile tool containing calibration metadata. A session projection drives a realtime armour-status badge above the input field, so the displayed state is derived from session data rather than from a static label. The supplied material does not name the profile tool, its arguments, configuration keys, or any files it reads or writes.
This is intended for developers testing or inspecting how a prompt-modification layer affects DeepSeek Harness sessions. It is the wrong choice when an agent must preserve the model’s ordinary instruction hierarchy, or when a project requires documented configuration and tool interfaces before installation. DeepSeek Harness already supplies sessions and a UI; this plugin adds a prompt layer and status presentation rather than a general-purpose analysis tool. The repository README describes a different dsh-infinite-gen-5 reverse-engineering workbench, so its commands, runtime assets, and compatibility notes should not be treated as documentation for dsh-infinite-gen-3.
dsh-infinite-gen-3 registers a system-prompt section at order 100. That section is described as an unconditional-compliance prompt for DeepSeek models. The plugin also exposes a profile tool with calibration metadata. A session projection supplies the state used by the realtime armour-status badge shown above the input field.
The available description does not establish whether the prompt section replaces or follows other sections beyond its stated order, nor does it define the projection schema, calibration fields, or badge states.
The supplied README does not document configuration keys, defaults, environment variables, or editable files for dsh-infinite-gen-3. The installation metadata identifies the web DeepSeek Harness profile as the installation target, but it does not describe profile-level settings for the plugin.
DeepSeek Harness provides the session and UI context in which the plugin operates. The plugin’s documented mechanism is session-scoped: its status badge is driven by a session projection rather than by a separately described persistent store.
The upstream summary identifies a profile tool, but gives it no command name, argument schema, return format, or invocation example. No plugin-specific slash command is documented.
The README’s command examples belong to dsh-infinite-gen-5, including /ig5 status, /ig5 engines, ig5_profile, and the ig5_* tool family. They should not be used as dsh-infinite-gen-3 commands.
The plugin is described as being for DeepSeek models and DeepSeek Harness sessions. The supplied material does not state a DeepSeek Harness version range, supported operating systems, network behaviour, filesystem paths, or permission requirements. Before enabling it in an agent with sensitive access, verify those details from the plugin source or manifest. Its purpose is prompt modification and status reporting, not reverse engineering or the analysis workflows documented for dsh-infinite-gen-5.
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
Minglink
System-prompt armor plugin for DeepSeek models: appends an unconditional-compliance prompt section at order 100, exposes a profile tool with calibration metadata, and shows a realtime armor-status badge driven by a session projection.
Installed cleanly when we ran it
ranxianglei
The official billion-context plugin: a context-compression plugin for small context windows (a 100K context is enough), token savings (5x fewer tokens), and month-long single sessions (billions of tokens).
Installed cleanly when we ran it
qiz029
A DeepSeek coding agent harness: persistent shell, Ultra subagents, auto approval, Chrome MCP and session telemetry
Did not install when we ran it
liangmianya
Visual, non-linear conversation workspace for DeepSeek Harness — sessions, follow-ups and branches become a browsable conversation map.
Installed cleanly when we ran it
Nwflower
Import full-fidelity chat histories from 13 coding agents (Claude Code, Codex, ChatGPT, Cursor, Gemini, opencode, and more) as resumable DeepSeek Harness sessions, with reverse export back to Claude Code.
Installed cleanly when we ran it
Totoro-qaq
Moves an existing DSH session to another agent preset through a previewable five-section handoff, preserving the source session and either pausing the target for confirmation or continuing immediately.
Installed cleanly when we ran it
