agentleFS
jsathjsath/claude-skills/skills/business/new-hire/SKILL.md

new-hire

Skill14 starsChanged 35 days agoMITView on GitHub

Onboards a new team member from the team's own documents with five agents: a two-week day-by-day training plan, a docs-only Q&A Project that says \"ask your manager\" when the docs are silent, end-of-day quiz questions for a Google Form, a gap list of what the docs could not answer, and a Friday progress check from the quiz and checklist Sheet. Use when a manager is onboarding someone, wants a training plan, a docs Q&A setup, daily quizzes, a list of doc gaps, or says \"build training\", \"quiz day 3\", \"find gaps\" or \"friday progress\". It never messages the new hire itself.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    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 three rounds, accept "skip", then show the PROFILE back for a yes.

    line 23
  2. Distinctive:

    Start date, working days, who else they can ask.

    line 26
  3. Distinctive:

    Read the agent file before running that step.

    line 49
  4. Distinctive:

    Never message, email or post to the new hire or the team.

    line 70
  5. Distinctive:

    No passwords, keys, pay or other people's personal data in any output or uploaded doc.

    line 72

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

New Hire

A new hire learns from your documents, answers a few quiz questions each day, and you get a Friday view of who is on track. You stop repeating the same walkthrough every hire.

Where it runs and what it connects to

PieceWhat it really doesNeeds
Claude Project with your docsThe new hire asks questions there; answers come only from the docsSharing a Project with another person needs a Team or Enterprise plan. Otherwise the hire builds their own Project from the same docs and the same instructions
Google Drive connectorReads your docs and the quiz SheetOptional. Without it, attach files or export the Sheet as CSV
Google Form + SheetHolds quizzes and checklist ticksYou create the Form from the questions Claude writes; Claude cannot create or send Forms
Claude CodeSame agents, files kept in onboarding/; friday progress can run as a scheduled taskOptional

Works in Claude.ai and Claude Code. The Friday check runs when you run it, unless you set up a scheduled task yourself.

Setup questions (start)

Ask in three rounds, accept "skip", then show the PROFILE back for a yes.

  1. Role, and what they should do alone by the end of week two.
  2. Start date, working days, who else they can ask.
  3. The docs: SOPs, process notes, FAQ, tool guides (attach, Drive, or Project).
  4. Tools they will use and who grants access.
  5. What new hires always get wrong.
  6. Team or Enterprise plan (can the Q&A Project be shared)?
  7. Google Form + Sheet, or CSV export? Sheet name.
  8. Pass mark (default 4 of 5).
  9. "Behind" rule (default 2+ days of checklist items not done).
  10. Anything that must never appear in training (client names, pay, disputes).

Commands

CommandRuns
startsetup questions, writes PROFILE
build trainingagent 1, then agent 3 for each of the 10 days
qa setupagent 2 setup mode: Project instructions to paste
quiz day <N>agent 3 for one day
find gapsagent 4
friday progressagent 4, then agent 5

Agents

Read the agent file before running that step; it is the only home for that role's detailed rules.

  1. agents/01-training-builder.md: ten-day plan and checklist, every item tied to a doc.
  2. agents/02-question-answerer.md: docs-only Q&A instructions and answer behaviour.
  3. agents/03-quiz-writer.md: 3 to 5 questions per day plus answer key, paste-ready for Google Forms.
  4. agents/04-gap-finder.md: unanswered questions and low-scoring quiz items become doc fixes.
  5. agents/05-progress-check.md: Friday status per hire from the Sheet, with row evidence.

Shared files

Claude Code: files in onboarding/. Claude.ai: print the updated table at the end of the run so the manager can save it to the Project.

  • PROFILE: setup answers.
  • PLAN: ten-day plan + checklist, with version and date.
  • QUIZZES: questions and answer keys per day.
  • GAPS: open gaps and fix status.
  • PROGRESS: one row per hire per Friday.
  • APPROVAL QUEUE: drafts for the hire or team (M1, M2...), sent by the manager only.

Safety rules

  • Never message, email or post to the new hire or the team. Anything for them is an M draft in the APPROVAL QUEUE.
  • Docs only: training, answers and quizzes use the documents or what the manager says in chat. Missing facts become GAPS, never guesses.
  • No passwords, keys, pay or other people's personal data in any output or uploaded doc.
  • Tell hires that quiz answers and checklist ticks are seen by their manager.
  • Progress is a coaching aid: every status cites Sheet rows. Never recommend firing, pay or probation outcomes.

End-of-run output

NEW HIRE: friday progress, week of 2026-01-05
4 Gap Finder     6 open questions grouped into 3 gaps, 1 quiz item under 60%
5 Progress       Sam Example ON TRACK, Alex Sample SLIGHTLY BEHIND (Day 4 quiz 2/5)
APPROVAL QUEUE   M1 to Alex Sample: offer 15 min on the refund doc (draft)
NEEDS YOU        GAP-2 "Who approves discounts over 10%?" no doc covers it
Missing data     Day 5 rows absent; Form may not be linked to the Sheet
Nothing has been sent to the new hire or the team.

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.