Introduction¶

The philosophy of DSH is “everything is a plugin”. For long, multi-step tasks like job hunting, the common approach is to write a long prompt and stack all requirements together. The result is unreproducible judgment criteria, unreviewable processes, and difficulty for users to intervene. dsh-career-skills takes a different approach: breaking job hunting down into four small skills—job-fit, resume-tailor, interview-prep, plus a career router—left to DSH to call on demand.

Below is an introduction to this plugin: who maintains it, what it does, how to install it, how to use it, and which boundaries you need to know beforehand.

What is it¶

dsh-career-skills is a DSH plugin released by maintainer lengrufeng2008, categorized as a workflow, derived from the career-ops methodology. It is positioned as four skills plus a router for job hunting. It does not scan job boards, does not auto-apply, the input is the job description (JD) or link pasted by the user, and the output is drafts to be reviewed by the user.

Notably, it is deliberately lightweight: no Node/Python scripts, no Playwright, no need for npx init, and no need to manually edit configuration files. The plugin uses DSH’s read, write, web_fetch, and ask_user_question tools directly to complete the workflow. It prioritizes local operations, does not auto-commit, does not send emails, and does not collect telemetry data. Every output stays in the draft phase, to be used only after user confirmation.

Core Functions¶

Four skills, the first three solve specific problems each, and the fourth handles triage.

1. job-fit: Determines whether to apply for a position. First, it checks the position’s activity and hard dealbreakers, then gives a score of 1–5, checks if essential requirements are covered, if the level and salary range match, and flags red flags.
2. resume-tailor: Rewrites the CV, resume, or cover letter into a version tailored to a specific JD. It follows ATS rules, maps experience to job requirements as evidence, and does not fabricate information where it is lacking, directly marking it as [NEED: ...].
3. interview-prep: Interview preparation. Includes company background research, STAR+R story bank, mapping predicted questions to stories, a list of questions to ask the interviewer, and a seven-day plan.
4. career: A router. When the user gives a request, it determines which of the above skills to use, preventing the user from not knowing which to call.

Differences from Direct Porting¶

The methodology for this comes from career-ops (MIT License). The author explicitly states that this is not a copy of the career-ops pipeline of over 600 scripts, but a native port to DSH:

• Four skills replace a complete CLI project; the routing mechanism keeps the skill directory within a small scope.
• Does not rely on Node/Python scripts or Playwright.
• Local-first, no auto-commit, no email sending, no telemetry.
• All outputs are drafts for user review.

Methodological inspiration and code attribution are explained in NOTICE.md. No career-ops code was copied; the skill text is original and natively adapted for DSH.

Installation and Activation¶

Installation depends on whether you are using a DSH bundle or directly copying the skill directory. Refer to the official README for installation commands.

If using a DSH bundle, you can install it via the plugin manager after publishing to npm:

dsh plugin --profile web add dsh-career-skills

When checking out locally, you can add the entire folder via Plugins → Add plugin, or use the plugin manager.

You can also copy the skills to DSH’s skill root directory:

./install.sh

Or manually copy:

mkdir -p "${DSH_HOME:-$HOME/.dsh}/skills"
cp -R skills/* "${DSH_HOME:-$HOME/.dsh}/skills/"

The difference between the two steps is that the former runs the plugin package’s installation script, while the latter directly overwrites the skill directory, suitable for those who want to see exactly what is installed.

Typical Usage¶

After installation, just ask in a normal conversation; no special commands are needed:

• “Evaluate this job against my CV: “
• “Tailor my CV to this role: “
• “I have an interview at for ; prepare me.”

You can also start with the router, letting the plugin decide which skill to use:

• “Use the career skill to decide which skill fits this.”

To allow the skills to read the necessary context, you need to place two files beforehand:

1. Place cv.md in the workspace and write the current resume content in it.
2. Write the target profile in .agents/career-profile.md, including target position, location, salary floor, and hard dealbreakers.

If either of these files is missing, the skills will ask follow-up questions via ask_user_question or mark [NEED: ...] where information is missing, rather than filling it in out of thin air.

Use Cases and Notes¶

It suits two types of people: developers who want to structure the job hunting process within DSH, and those who want resume rewriting and interview preparation to become reviewable drafts instead of one-off large prompts. It is categorized as a workflow, and the four skills run with the permissions of the current DSH process. It is recommended to read the source code and LICENSE before installation.

Its boundaries are also clearly stated:

• Does not scan job boards or auto-apply; requires the user to paste the job description or URL.
• Does not generate PDF unless the system has the conversion tool installed.
• It does not know the user’s salary history or private information and will ask follow-up questions or mark [NEED: ...] accordingly.
• It is not a headhunter and does not guarantee results.

The license is MIT; see LICENSE and NOTICE in the repository.

Conclusion¶

The value of dsh-career-skills lies not in its many features, but in breaking down job hunting into small, reviewable, and routable units, while maintaining local-first, script-free, and auto-commit-free operations. For those already using DSH for automation, it is a reference implementation that can be installed directly or adapted.

Directory Page: https://www.skillhub.cn/plugins/lengrufeng2008/dsh-career-skills

GitHub: https://github.com/lengrufeng2008/dsh-career-skills