setup-openings
Skill7 starsChanged 3 days agoMITView on GitHub
Set up a new or empty Openings workspace for its owner - "set up my Openings workspace", "get me started". Checks the connection and local tools, builds the Profile and career entries from a CV, records Preferences from a short interview, creates CV and letter templates, adds first sources and runs a first triage.
In this repository8 other files it gives its agents.
- AGENTS.mdAGENTS.md
- maintain-careerSkill
- manage-sourcesSkill
- openingsSkill
- prepare-applicationsSkill
- prepare-interviewSkill
- track-applicationsSkill
- triage-jobsSkill
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 3 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 11
Read openings first. Do the steps in order.
- Very specific: line 16
Call workspace_context. If it fails, the owner needs a running server (the repository's getting-started guide) and an Agent token from …
- Distinctive: line 19
When it names you as the owner, suggest an Agent token: it can do all of this work.
- Very specific: line 21
Check you have a shell with curl for uploads, and pdflatex or another TeX engine for PDFs.
- Distinctive: line 27
Ask for a CV, LinkedIn export or a description of their work, and follow maintain-career to build the Profile, entries and highlights.
- Very specific: line 50
Use the owner's own TeX CV and letter when they have them.
- Distinctive: line 57
Follow manage-sources: searches for the target roles and places, and boards of companies they name, previewed before saving.
- Very specific: line 69
If the owner's client loads skills from a folder, offer to save these skills there: read each with skills_get and write it, with its …
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Set up your workspace
Read openings first. Do the steps in order; each uses what the previous one recorded. Skip what is already done.
1. Check the connection
- Call
workspace_context. If it fails, the owner needs a running server (the repository's getting-started guide) and an Agent token from Settings › API & agents in your client's environment. Never ask for the token in chat. When it names you as the owner, suggest an Agent token: it can do all of this work. - Check you have a shell with
curlfor uploads, andpdflatexor another TeX engine for PDFs. If TeX is missing, offer to install a small distribution (such as TinyTeX) with the owner's agreement.
2. Career entries
Ask for a CV, LinkedIn export or a description of their work, and follow maintain-career to build the Profile, entries and highlights. Ask about gaps that matter: dates, own role, results.
3. Preferences
Interview the owner briefly, a few questions at a time:
- What I'm looking for: roles, seniority, places, remote, company types;
- What I rule out: dealbreakers;
- Facts about me: work permit, availability, notice, salary, languages;
- Writing rules: tone, length, spelling, words to avoid, for the CV, the letter or both;
- Working with me: how they want you to work, such as asking before large dismissals.
Record each answer as a separate item in their words with
preferences_update. Set the languages they speak and, if they say, the days
before an application counts as unanswered, with settings_update. Tell them
they can edit everything under Preferences.
4. Templates
Use the owner's own TeX CV and letter when they have them; otherwise write a
clean one-page CV and a letter template. Compile each with real content,
check the PDF, and save it with templates_save, with guidance on what to
tailor and limits such as pages.
5. First sources
Follow manage-sources: searches for the target roles and places, and boards of companies they name, previewed before saving.
6. First results
Once jobs arrive, run triage-jobs, and prepare one shortlisted application with prepare-applications so the owner sees a real result.
Skills on disk
If the owner's client loads skills from a folder, offer to save these skills
there: read each with skills_get and write it, with its references, to that
folder (for example .claude/skills/<name>/).
Report back
In chat: what is set up (entries, Preferences items, templates, sources); the open questions about their career; and their part from now on: review Ready applications, send them and tell you, or mark them Applied, when they do.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

