agentleFS
gladosAagmangladosAagman/linkedin-agent-skill/skills/li-jobsearch/SKILL.md

li-jobsearch

Skill1 starsChanged 5 days agoMITView on GitHub

Run a job search on LinkedIn the way hiring actually works - a profile that shows up in recruiter search, a target list of companies and people, outreach to hiring managers and referrers instead of mass applying, and replies to recruiters that get the facts. Use when the user says "I'm job hunting", "find me a job", "optimize my profile for recruiters", "open to work", "message a hiring manager", "ask for a referral", or "how do I reply to this recruiter".

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ask, in one batched question:

    line 22
  2. Distinctive:

    A referral from someone they know: Ask for the specific thing: "Would you be comfortable referring me for {role}?

    line 57
  3. Distinctive:

    When a recruiter messages, ask the three things the message left out: the compensation band, the level, and the location policy.

    line 71
  4. Distinctive:

    Never apply, message, connect or post on the user's behalf.

    line 85
  5. Distinctive:

    Never inflate a title, a date or a result.

    line 87
  6. Distinctive:

    Never fabricate a mutual connection or a referral.

    line 89

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

li-jobsearch

Mass applying is the default and it is the worst-performing option. One-click applications land in a pile of hundreds. A referral or a message to the hiring manager lands in front of a person. This skill builds the second kind of search and keeps the user out of the first.

Before anything, get the target

Ask, in one batched question:

  1. The role, in the words job postings use. Two titles at most.
  2. The level, the location or remote, and the floor on compensation.
  3. Ten to twenty companies they would actually say yes to.
  4. Proof: two or three outcomes from recent roles, with numbers. The skill never invents one. Without them, every draft comes back with {{your number}} in it.

1. Be findable

Recruiters search by title, skills and location. Run /li-profile with the target role in mind, then check:

  • Headline: the target title, word for word, then what changes when they hire you. ../li-post/count.py --kind headline for length.
  • About: the first two lines say the role and the proof.
  • Experience: outcomes with numbers, in the job posting's vocabulary.
  • Skills: the ten that appear most in the target postings, in that wording.
  • Open to Work: say plainly what each setting shows. "Recruiters only" is visible to people using LinkedIn Recruiter, and LinkedIn says it takes steps to hide it from recruiters at the user's own company but cannot guarantee it. The public photo frame is visible to everyone. The user decides.

2. The target list

For each company: the likely hiring manager (the role the job reports to), one person in the team, and anyone the user already knows there, first or second degree. Ask the user to find these names on LinkedIn themselves. Do not search or scrape on their behalf.

3. Outreach, in this order

  • A referral from someone they know. Ask for the specific thing: "Would you be comfortable referring me for {role}? Here is the posting and three lines on why I fit." Make it easy to say yes and easy to say no.
  • The hiring manager. A note or message with one specific reference to their team's work, one line of proof, one small ask. Write it with /li-dm and check it with count.py --kind invite.
  • The application, after the message, not instead of it. Say in the message that they have applied.

Ten good messages a week beats a hundred applications. Never more than twenty invites a day.

4. Recruiter replies

When a recruiter messages, ask the three things the message left out: the compensation band, the level, and the location policy. If the role is wrong, one line: not a fit, what would be, and a referral if the user knows someone. /li-inbox triages a full inbox.

5. Posting while searching

Optional. If the user wants to, one post a week about the work they do, not about the search, through /li-post. A post about looking for work is fine once, written plainly, with the role and what they are good at. It is not a weekly format.

Never

  • Never apply, message, connect or post on the user's behalf. Automated outreach violates LinkedIn's User Agreement.
  • Never inflate a title, a date or a result. Background checks exist, and so does the hiring manager's memory.
  • Never fabricate a mutual connection or a referral.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.