Install verification: Ecosystem app: not installed via dsh plugin add, so the L1-L5 plugin install-verification ladder does not apply; the security scan still runs.Install verification details →
Data as of (UTC)
What it doesAI
Electron-based WeChat-style desktop client for DeepSeek Harness with chat, approvals, and model switching.
- WeChat-style three-pane UI (icon bar, session tree, chat window)
- Fully custom chat: Markdown, streaming, optimistic rendering
- Auto-discovers local dsh instance, manages connections
Capability tags are AI-inferred or rule-extracted from the README (see badge) — vocabulary-limited, not install-verified. AI = inferred; README = literal keyword from the readme.
Installation
Ecosystem apps are not installable via dsh plugin add: desktop / web shells install or deploy through their own release channel (GitHub Releases or the project site).
The level reads L5 · Ecosystem plugin / L5 · Ecosystem app per the three-way classification (the L1-L5 ladder does not rate them); the security scan still runs.
Trends
Four lines, four questions: total stars, stars gained this week, cumulative downloads, downloads this week. The GitHub pair comes from real-read sampling — a rotating daily shard, so one plugin is re-read roughly every 5–14 days, sparser than npm; the npm pair has a value on every snapshot day. The two cumulative panels are smoothed curves with no sample markers; the two weekly panels keep their raw bars and add a 4-week moving-average trend line that strips single-week noise down to the direction of travel.
9 samples (roughly every 7 days), 5 of them reconstructed (dashed runs); a smoothed curve through every sample — not a daily series
09-18 → 11-05 · 9 points
Bars = the delta between two samples 5–14 days apart (negative points down, faint bars contain reconstructed points); line = 4-week moving-average trend
09-28 → 11-05 · 8 points
Reconstructed backwards from the current total using daily downloads (gaps are not interpolated)
Only 0 points (< 3): no chart yet
no data yet
Bars = the snapshot’s last-week rolling window (official npm + npmmirror combined); line = 4-week moving-average trend — the shape of weekly demand, not a user count
Only 0 points (< 3): no chart yet
no data yet
Verification & Compatibility◼ Ecosystem app
Grouped by plugin version: each card lists the dsh versions actually tested against that artifact version and their install verdicts (L5 wins). Untested combinations are omitted — absence means untested, never assumed compatible. L1–L3 are a plugin-level static archive (independent of the dsh version — the same verdict on every dsh row): they sit in each row's level grid next to L4/L5, marked "static", while their evidence is stored once per plugin inside that row's "checks & evidence" instead of being duplicated per version.
External credentials (plugin-level): none detected
| plugin \ dsh | dsh 0.1.5-rc.2 | dsh 0.1.3-alpha.2 | dsh 0.1.3-alpha.1 | dsh 0.1.2-rc.1 | dsh 0.1.2-alpha.1 |
|---|---|---|---|---|---|
| v0.1.0 | ◼ | ◼ | ◼ | ◼ | · |
v0.1.0github◼ 4 ecosystem app2026-09-11
https://github.com/MerlinShieh/deepseek-herness-desktop
◼dsh 0.1.5-rc.22026-09-11Ecosystem app
- • Node.js: v24.14.0
- • DSH: 0.1.5-rc.2
- • Plugin artifact: github · https://github.com/MerlinShieh/deepseek-herness-desktop#v0.1.0
ID: deepseek-herness-desktop@v0.1.0@dsh0.1.5-rc.2 · profile: l4-sandbox
Issued: 2026-09-11 · expires: 2026-09-18
View checks & evidence
not present in web-profile loader inventory
L1–L3 are plugin-level static conclusions (baseline, identical across dsh versions): evidence is stored once per plugin.
◼dsh 0.1.3-alpha.22026-09-08Ecosystem app
- • Node.js: v24.14.0
- • DSH: 0.1.3-alpha.2
- • Plugin artifact: github · https://github.com/MerlinShieh/deepseek-herness-desktop#v0.1.0
ID: deepseek-herness-desktop@v0.1.0@dsh0.1.3-alpha.2 · profile: l4-sandbox
Issued: 2026-09-08 · expires: 2026-09-15
View checks & evidence
not present in web-profile loader inventory
L1–L3 are plugin-level static conclusions (baseline, identical across dsh versions): evidence is stored once per plugin.
◼dsh 0.1.3-alpha.12026-09-07Ecosystem app
- • Node.js: v24.14.0
- • DSH: 0.1.3-alpha.1
- • Plugin artifact: github · https://github.com/MerlinShieh/deepseek-herness-desktop#v0.1.0
ID: deepseek-herness-desktop@v0.1.0@dsh0.1.3-alpha.1 · profile: l4-sandbox
Issued: 2026-09-07 · expires: 2026-09-14
View checks & evidence
not present in web-profile loader inventory
L1–L3 are plugin-level static conclusions (baseline, identical across dsh versions): evidence is stored once per plugin.
◼dsh 0.1.2-rc.12026-09-05Ecosystem app
- • Node.js: v24.14.0
- • DSH: 0.1.2-rc.1
- • Plugin artifact: github · https://github.com/MerlinShieh/deepseek-herness-desktop#v0.1.0
ID: deepseek-herness-desktop@v0.1.0@dsh0.1.2-rc.1 · profile: l3-fullscan
Issued: 2026-09-05 · expires: 2026-09-12
View checks & evidence
ecosystem plugin: no dsh.bundle declaration
L1–L3 are plugin-level static conclusions (baseline, identical across dsh versions): evidence is stored once per plugin.
Only the latest 3 published versions are shown; source-snapshot (commit hash) groups, unstated groups and older versions are omitted — the full archive stays in the data layer (snapshot/unstated groups appear only when a plugin has no published version).
Author badges
Embed the dsh.so badges in your README to show this plugin's automated security-scan and install-test status:
`[](https://www.dsh.so/artifact/deepseek-herness-desktop/?ref=badge)`dsh.so is an independent community project, not affiliated with or endorsed by DeepSeek AI. Badges are maintained by dsh.so; embed links carry a ?ref= attribution parameter.
Security Report
Two automated layers, neither a manual review: a static scan (per-plugin rule checks) and a model-driven dynamic security audit that covers only the subset static-flagged at warn or above and never executes plugin code.
Static low; dynamic high (authoritative when present) → overall high.
v0.1.0v0.1.0git2026-11-17v0.1.0same versionclean · gitVet is an AST static scan (npm artifact or git source).
show static findings (3)
- criticalshellExec · shell
src/main/index.ts:261exec(`start "" "${connectedUrl}"`) - warningchildProcess · code-exec
src/main/dsh-service.ts:137const proc = spawn(command, args, { - warningnetIp · network
src/main/dsh-service.ts:43const url = `http://127.0.0.1:${port}`
show 4 audit findings
- highShell command injection: cmd.exe string built from renderer-supplied dsh URL
src/main/index.ts:261connectedUrl comes from the IPC handler ConnectInstance(_e, url) / config lastUrl and is only gated by one successful POST /api/host.describe, so it can be any attacker-chosen URL (e.g. http://evil.tld/" & calc & " ; fetch percent-encodes the quotes but the raw string is kept for the shell). If shell.openExternal rejects (the very case this fallback exists for), cmd.exe executes the URL as a command line and the injected `& calc &` runs arbitrary commands in the user's context.
- highNo will-navigate guard: remote page loaded in the shell window keeps the preload IPC bridge
src/main/window.ts:34The preload at src/preload/shell.ts (webPreferences.preload, line 23) runs on every navigation of that webContents and exposes the whole contextBridge API. Chat messages are rendered as markdown anchors (renderMarkdown -> DOMPurify.sanitize keeps <a href>), so a link produced by the agent/model navigates the window to an attacker origin, which then has window.dshDesktop: connectInstance(arbitrary URL), chatSend, discoverInstances, openFullUI (the exec sink above) - a full renderer-to-main privilege escalation.
- lowChromium sandbox globally disabled for a window that renders untrusted model output
src/main/index.ts:28The OS sandbox and the renderer sandbox are both off, so any renderer-side code execution (e.g. a DOMPurify bypass in the markdown renderer) immediately becomes native code execution with no sandbox escape required. The comments say this is a workaround for VM/RDP and Windows sandbox error_code=40; no reachable sanitizer bypass was found.
- lowUser chat text and attachment names written to a plaintext log file
src/main/index.ts:328Prompt content (first 60 chars of every message, plus attachment file names, session ids and connected URLs) is appended in plaintext to userData/dsh-desktop.log (log.ts appendFileSync) and to stdout. Anyone with read access to the profile (other local apps, backups, support bundles) can recover fragments of the user's conversations and file names.
Verified the static warning: main process runs a shell command built from a renderer-supplied URL, and the window can navigate to remote pages that keep the preload IPC bridge.
Rule (2026-10-28): when a dynamic conclusion exists it is authoritative (it may raise or lower the level); the static scan tier (vet primary, dsh backstop) applies only when there is none (not queued / audit failed / withdrawn). The audit never writes the exemption list (exemptions are manual only); a wrong conclusion is withdrawn by human review and the tier falls back to the static scan tier.
Disclaimer: the static scan and the dynamic security audit are both automated analysis, not a security guarantee. Always review what you install.
Activity
Last commit 2026-08-15 · activity: Inactive
• Repo created: 2026-08-15
• Stars: ★ 1 · Forks: 0
• Health: Inactive — no commit in 90+ days
Source
Author README
View on GitHub ↗The author's original README from the repository — fetched live, unedited (the English page loads the default README).
Frequently asked questions
What is deepseek-herness-desktop?
deepseek-herness-desktop is an ecosystem app listed on dsh.so, not a DSH plugin package: it cannot be installed with dsh plugin add and ships through the project's own release channel.
Is dsh.so or this plugin page official?
dsh.so is an independent community registry and is not affiliated with, endorsed by, or sponsored by DeepSeek AI.
How do I install deepseek-herness-desktop?
This is an ecosystem app, not a plugin installable via dsh plugin add: desktop / web shells install or deploy through their own release channel. The security scan still runs, but the L1-L5 plugin install ladder does not rate it.
How does the security scan relate to a manual audit?
Both layers are automated, not a manual audit. The static scan is heuristic analysis of the repository source, so it can produce false positives and miss context. The model-driven dynamic security audit only covers plugins static-flagged at warn or above, never executes plugin code, and its conclusion is authoritative when there is one (it may raise or lower the static tier); exemptions stay manual. Neither layer is an endorsement: review the repository and the permissions it asks for before installing, together with the verification level and health signals.
How fresh is the data on this page?
The data on this page is as of 2026-11-17 (UTC) — the most recent valid timestamp among the registry refresh, install verification, security scan and dynamic audit. Build time is never used as a data date.
