Skills/npm/dsh-mcp-panel

dsh-mcp-panel@0.6.34

VERIFIED· install only

Publisher claims: "MCP management console for the official DeepSeek Harness MCP client: /mcp command with health diagnostics and pipeline trial calls, a Settings MCP tab with server CRUD (append-only profile-patch fragments, approval-gated writes with automatic backups), a " Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry npm (MCP server)Publisher perrylink · publishing for 101 daysInstalls 11,542Last detonated 1 h ago
$claude mcp add dsh-mcp-panel -- npx -y dsh-mcp-panel
WatchReport an issue

Coverage

Bundled scripts ran with no arguments; no agent task scenario for this bundle.
Tested
4 of 4 bundled scripts run with no arguments, with decoy credentials in the environment · every file read, file written and host contacted by what ran
Not tested
use of the skill by an agent

What changed since 0.6.29

vet-skills diff dsh-mcp-panel 0.6.29 0.6.34

This version reaches for something 0.6.29 did not: 1 new credential variable (CREDENTIAL_KEY_SOURCE). Compared from the two recorded runs, not from source.

New credential variables
CREDENTIAL_KEY_SOURCE

The guard can hold an update like this until someone approves it (runtime.on_change: hold). A change that only shows with a real credential, or on a path the run did not reach, is not in a trace and so not in this comparison.

Is [email protected] safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (2 scenarios; bundled scripts ran with no arguments; no agent task scenario for this bundle) on 2026-11-22 and scored it 80/100 (Verified). In what ran, nothing we could detect went wrong: no undeclared network egress, no credential reads, no persistent instruction writes. What it does with a real credential was not observed; the static findings and the publisher carry the rest of the score. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 1 s · 2 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
0.0sprocessbundled script run in the sandbox with no arguments
0.1sprocessexited 1: Node.js v22.23.5
0.1sprocessbundled script run in the sandbox with no arguments
0.2sprocessexited 1: Node.js v22.23.5
0.2sprocessbundled script run in the sandbox with no arguments · not named in the instructions
0.3sprocessbundled script run in the sandbox with no arguments
0.5sprocess4 of 4 bundled scripts run with no arguments
0.5sprocessscript backend runs setup only
0.5sprocessscript backend runs setup only
0.5sprocessno install scripts to re-run

Static findings

HIGH
Collapsed <details> block carrying instructionsREADME.en.md:253 · Collapsed <details> block carrying instructions<summary>Is the connection status live? What is shown when nothing is observed?</summary> Status comes from the official client's `mcp/status` seam (not yet shiFix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
MEDIUM
Process kill commandsrc/service.ts:400 · Process kill command — agent can terminate processes* - `writeEnabled: false` rejects every write up front (kill switch).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/probe.ts:96 · Process kill command — agent can terminate processes* @param signal - caller-owned abort (job kill or timeout).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandlib/types/probe.d.ts:61 · Process kill command — agent can terminate processes* @param signal - caller-owned abort (job kill or timeout).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandlib/types/service.d.ts:145 · Process kill command — agent can terminate processes* - `writeEnabled: false` rejects every write up front (kill switch).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandlib/index.js:2579 · Process kill command — agent can terminate processes* - `writeEnabled: false` rejects every write up front (kill switch).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Collapsed <details> block carrying instructionsREADME.en.md:267 · Collapsed <details> block carrying instructions<summary>Does panel content reach model context? Can secrets or headers leak?</summary> No. Panel content (status cards, CRUD, the trial console, the Resources Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
HIGH
Image alt text carrying instructionsREADME.en.md:126 · Image alt text carrying instructionsdsh-mcp-panel: real terminal runFix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
MEDIUM
Process kill commandsrc/probe.ts:162 · Process kill command — agent can terminate processes* @param signal - caller-owned abort (job kill or timeout).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/config.ts:48 · Process kill command — agent can terminate processes/** Whether profile-patch writes are allowed at all (kill switch; default true). */Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandlib/types/config.d.ts:35 · Process kill command — agent can terminate processes/** Whether profile-patch writes are allowed at all (kill switch; default true). */Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/client/ServerEditor.tsx:28 · Process kill command — agent can terminate processes/** Whether profile writes are enabled (kill switch). */Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Auto-update or dynamic fetch configurationsrc/command.ts:97 · Auto-update or dynamic fetch configuration — agent may load new code at runtimepatchReloadPath: 'The web surface hot-reloads cordis.patch.yml edits; other surfaces restart. This command never edits your config.',Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commandlib/index.js:924 · Process kill command — agent can terminate processes* @param signal - caller-owned abort (job kill or timeout).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandlib/index.js:861 · Process kill command — agent can terminate processes* @param signal - caller-owned abort (job kill or timeout).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/wire.ts:201 · Process kill command — agent can terminate processes/** Whether profile-patch writes are allowed at all (kill switch). */Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandlib/types/wire.d.ts:187 · Process kill command — agent can terminate processes/** Whether profile-patch writes are allowed at all (kill switch). */Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/service.ts:409 · Process kill command — agent can terminate processesthrow new Error('dsh-mcp-panel: profile writes are disabled (config.writeEnabled: false) — flip the kill switch or edit the file by hand')Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Collapsed <details> block carrying instructionsREADME.md:267 · Collapsed <details> block carrying instructions<summary>面板内容会进入模型上下文吗?密钥和 header 会不会泄露?</summary> 不会。面板内容(状态卡、CRUD、试用台、Resources 浏览)只显示在设置页,永不进入模型上下文;只有 `/mcp` 命令输出对模型可读、并可由会话日志重建。渲染前会打码 URL 查询凭据、userinfo 密码Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
MEDIUM
Process kill commandlib/types/probe.d.ts:77 · Process kill command — agent can terminate processes* @param signal - caller-owned abort (job kill or timeout).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Auto-update or dynamic fetch configurationlib/index.js:325 · Auto-update or dynamic fetch configuration — agent may load new code at runtimepatchReloadPath: "The web surface hot-reloads cordis.patch.yml edits; other surfaces restart. This command never edits your config.",Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commandlib/index.js:2587 · Process kill command — agent can terminate processesif (!this.config.writeEnabled) throw new Error("dsh-mcp-panel: profile writes are disabled (config.writeEnabled: false) — flip the kill switch or edit the file by hand");Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/config.ts:79 · Process kill command — agent can terminate processes/** Whether profile-patch writes are allowed at all (kill switch). Default true. */Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandlib/types/config.d.ts:63 · Process kill command — agent can terminate processes/** Whether profile-patch writes are allowed at all (kill switch). Default true. */Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Collapsed <details> block carrying instructionsREADME.en.md:260 · Collapsed <details> block carrying instructions<summary>Does the console rewrite my cordis.patch.yml? What if a write goes wrong?</summary> It never rewrites the file. Writes are generated append-only fragmeFix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
MEDIUM
Process kill commandlib/types/client/ServerEditor.d.ts:17 · Process kill command — agent can terminate processes/** Whether profile writes are enabled (kill switch). */Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Collapsed <details> block carrying instructionsREADME.hi.md:267 · Collapsed <details> block carrying instructions<summary>क्या पैनल सामग्री मॉडल संदर्भ में जाती है? सीक्रेट या header लीक हो सकते हैं?</summary> नहीं। पैनल सामग्री (स्टेटस कार्ड, CRUD, परीक्षण कंसोल, ResourceFix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01

Cost to the agent

Instructions loaded
≈91 tok
1 file · 70 in bundle

Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.

Known vulnerabilities

No advisories in OSV for dsh-mcp-panel@0.6.34 as of 35 min ago. This is the package itself, not its dependencies.