linkedin-search-plan
Skill12 starsChanged 4 days agoMITView on GitHub
Translate a job-search brief into focused role and location queries, ordinary LinkedIn search links and a bounded Scout search proposal. Use when the user wants to plan or improve a LinkedIn job search, rather than evaluate a particular vacancy.
In this repository5 other files it gives its agents.
- company-briefSkill
- fit-reviewSkill
- job-researchSkill
- job-search-briefSkill
- shortlist-reviewSkill
The file
Files
LinkedIn Search Plan
Use the user's existing brief. If none exists, collect the target role and geography and only the missing preferences that change results. A CV is optional for query planning. Keep desired role titles separate from skills, exclusions and logistics; do not combine every preference into an overly narrow query.
Propose a small useful set of primary title queries and evidence-backed synonyms. Explain why an adjacent title is included. Keep location text unambiguous, such as United Kingdom rather than UK. Country-wide remote is a work mode, not proof of permission to work remotely from any country.
For manual LinkedIn browsing, generate ordinary job-search URLs using URL-encoded keywords and location:
https://www.linkedin.com/jobs/search/?keywords=Product%20Analyst&location=Berlin
These are navigation aids, not discovered vacancies. Prefer the user setting other filters in LinkedIn over claiming undocumented URL parameters are stable. Do not automate a browser, retrieve cookies or ask for LinkedIn credentials.
If Scout is connected, inspect get_scout_status and prepare_search schemas. Use prepare_search(brief, source) to validate and store the plan. Its current brief supports target_titles, locations, workplace, posted_within, easy_apply_only, max_results and max_charge_usd. Easy Apply defaults to false: ordinary research should include external-application roles too. Do not add unsupported filters such as company headcount or salary to the tool call; retain them as later review criteria.
Describe the source, search criteria, maximum results and run-charge cap in the conversation before any live search. Preparing the plan itself makes no paid call. Use the user's own configured Apify account; credentials are not tool arguments or chat content. When the user requested only a plan, stop at the plan. When the user already requested a bounded search, continue within that scope after making the caps visible; do not turn every routine step into another confirmation request.
For a demo, select source demo explicitly and describe the output as synthetic. Without Scout, provide the searches directly in chat and explain that the plan has not queried current listings. A plan, a search link and an observed vacancy are distinct outputs.
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.

