dsh-web is a plugin ecosystem that adds a web interface and optional tools to DeepSeek Harness, an AI development environment. It is used to provide features such as task boards, mobile remote control, SSH operations, image understanding, file and code panels, Git views, agent presets, and interchangeable skins through independently loaded plugins. The catalogue skills are plugins and related workflows for extending and operating the DSH Web environment.
Borrowing it
Nothing to install: this file belongs to zhu1090093659/dsh-web. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/zhu1090093659/dsh-web/dev/.agents/skills/dsh-web-code-review/SKILL.mdgit clone --depth 1 https://github.com/zhu1090093659/dsh-webWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/zhu1090093659/dsh-web/dsh-web-code-review)<a href="https://agentmods.dev/skills/zhu1090093659/dsh-web/dsh-web-code-review"><img src="https://agentmods.dev/badge/skills/zhu1090093659/dsh-web/dsh-web-code-review/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/zhu1090093659/dsh-web/dsh-web-code-review"><img src="https://agentmods.dev/badge/skills/zhu1090093659/dsh-web/dsh-web-code-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 20 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00046 | $0.00676 |
| Opus 5 | $0.00023 | $0.00338 |
| Sonnet 5 | $0.00009 | $0.00135 |
| Haiku 4.5 | $0.00005 | $0.00068 |
Grade A, and why
dsh-web-code-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 36 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reviewing dsh-web Changes
This skill is guidance, not a checklist substitute. Review the requested diff and enough surrounding code to establish actual behavior; do not infer correctness from a passing typecheck alone.
Establish the review frame
- Confirm the repository, requested base and head, and current worktree state. Do not mix unrelated dirty changes into the review scope.
- Read AGENTS.md, packages/AGENTS.md for package work, and the nearest package-specific instructions.
- Inspect the complete change surface: source, tests, manifests, generated artifacts, documentation, and the real host or client entry path.
Collaborative review
- Aa728848 is a collaborator with merge permission on
devand the auto-approver for renderer / Wallpaper Engine / WebGL PRs (routed by.github/workflows/auto-assign-pr-reviewers.ymlwith.github/pr-review-routes.json, see PR_TRIAGE.md). Those PRs normally skip our second review; do not review them unless the user explicitly asks. - Before reviewing or building on top of new code, sync his merged work first:
git fetch originand rebase ontoorigin/dev(git rebase origin/dev), then re-run the affected checks. His merged PRs are already ondev; never review or test against a stale localdev. - When the reviewed diff includes commits from his merged PRs, treat those commits as merged upstream code, not as part of the change under review.
Prioritize material risks
- Trace host, client, and shared changes through their actual consumers. Flag incompatible service assumptions, missing disposal, stale subscriptions, HMR regressions, and changed user-visible behavior without coverage.
- Enforce the SDK boundary: packages must not import or reference a DSH source checkout; browser bundles must preserve type-only SDK imports and allowed runtime dependencies.
- Check generated ownership. Changes to
shared/, aggregate manifests, skin registries, community indexes, or market assets need their corresponding generator and consistency check. - Treat credentials, remote execution, local filesystem access, profile writes, and token handling as security-sensitive. Confirm authorization, error behavior, cleanup, and documentation where the affected package requires them.
- For UI changes, inspect responsive layout, loading and error states, keyboard behavior where applicable, and actual rendered output. A component-only test does not prove the linked GUI integration works.
- Verify README pairs and documentation claims when public behavior, configuration, or safety semantics change.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 36 lines · 46 tokens per session scan A 2aaa10cb6da6
dsh-web-code-review is a skill published in the GitHub repository zhu1090093659/dsh-web (7,321 stars, last pushed today), licensed Apache-2.0. It adds 46 tokens to every session and 676 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
triage
Move issues and external PRs through a state machine of triage roles, categorise, verify, grill if needed, and write agent-ready briefs.
code-review
Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use…
improve-codebase-architecture
Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.
manage-taskboard
Manage work in the native DeepSeek Harness Taskboard with exact task ids and optimistic versions. Use when an Agent must inspect project work, claim an eligible todo, record progress or blockers, verify an implementation, submit it for human review, or release its own claim; also use when a human asks how to accept…
code-review-cn
A Chinese-language guide for reviewing code changes, pull requests, and diffs. It prioritizes correctness, resource cleanup, security, matching documentation, and tests before style.
refactor-safely
A safe code-refactoring guide for changing code structure while keeping its external behavior the same. Refactoring means reorganizing code without changing what it does.