Assisted job-application filler. An agent skill from extrasmall0/dear-hiring-manager.

MITAuto-check passedBusiness, Finance & HR

Install Apply

skills CLI
$ npx skills add extrasmall0/dear-hiring-manager --skill apply -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install extrasmall0/dear-hiring-manager apply --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/extrasmall0/dear-hiring-manager.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/apply .claude/skills/apply && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
apply
GitHub stars
111
Token cost
~2.6k tokens
SKILL.md length
1,350 words
Files
1
Skills in repo
7
Repo updated
First seen
Licence
MIT

At a glance

Assisted job-application filler. An agent skill from extrasmall0/dear-hiring-manager.

  • Works in 3 steps: Open the posting. Use the Playwright MCP… → Extract JD + company. Pull role title,… → Score fit — FIRST FILTER (hard gate,…
  • Whenever the user wants help applying to a specific job
  • SKILL.md covers Preconditions, Procedure, Rules and ATS notes (from dogfooding)
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Apply is an agent skill from extrasmall0/dear-hiring-manager. Assisted job-application filler. Given a job posting URL, open it in a browser via Playwright, extract the JD and company, score resume-to-JD fit, fill every form field from the user's profile and answer memory, flag anything uncertain, and STOP before Submit for human review. After the human submits, capture their edits back into answer memory. Use for /apply or whenever the user wants help applying to a specific job.

Its SKILL.md is about 2.6k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Business, Finance & HR, covering Job search and resumes, Browser testing and Resume and CV writing. It works with Playwright. The repository describes itself as: A Claude Code plugin that fills your job applications from your profile, drafts the cover letter, and leaves Submit to you. Locally. The licence is MIT.

When your agent uses it

  • Whenever the user wants help applying to a specific job
  • Tasks that involve Job search and resumes
  • Tasks that involve Browser testing

Example prompts

  • “/apply”

Workflow steps

3 steps, taken from the first numbered list in SKILL.md.

  1. Open the posting. Use the Playwright MCP tools to navigate to the URL. Take a page snapshot
  2. Extract JD + company. Pull role title, company, location, and the job description. Detect the
  3. Score fit — FIRST FILTER (hard gate, nothing filled before it passes). Read resume.* and the

What it can do on your machine

Read from SKILL.md and the folder at commit ff6c26b. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Apply loads about 2.6k tokens when it runs. Until then it costs about 107 tokens; SKILL.md has 1,350 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~107
When it runs · the whole SKILL.md, loaded when a task matches
~2.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from extrasmall0/dear-hiring-manager at commit ff6c26b, republished under its MIT licence (© extrasmall0). 1,350 words, ~2,581 tokens.

Download SKILL.mdSave it as .claude/skills/apply/SKILL.md (or your agent's skills folder).
name
apply
description
Assisted job-application filler. Given a job posting URL, open it in a browser via Playwright, extract the JD and company, score resume-to-JD fit, fill every form field from the user's profile and answer memory, flag anything uncertain, and STOP before Submit for human review. After the human submits, capture their edits back into answer memory. Use for /apply or whenever the user wants help applying to a specific job.

Apply (assisted, human owns Submit)

Fill one job application from a URL. You never click Submit. You fill, flag, and hand back a browser tab parked at the Submit button for the human to review, edit, and send.

Data lives in ~/.dear-hiring-manager/: profile.md, answers.md, applications.md, resume.*.

Preconditions

  • If ~/.dear-hiring-manager/profile.md is missing, stop and tell the user to run /dear-hiring-manager:onboard first.

Procedure

  1. Open the posting. Use the Playwright MCP tools to navigate to the URL. Take a page snapshot (accessibility tree / DOM) — do not rely on vision unless the DOM is unusable. If the URL redirects to the board root or an error page (e.g. ?error=true, only a job list / "Back to jobs", no job title or application form), the posting is closed or expired — stop and tell the user; do not proceed.

  2. Extract JD + company. Pull role title, company, location, and the job description. Detect the ATS from the URL/DOM (greenhouse, lever, ashby, workday, other). Append a row to applications.md (create from ${CLAUDE_PLUGIN_ROOT}/templates/applications.template.md if missing) with status=in-progress (fit is filled in the next step; status advances as the run proceeds): date | company | role | url | ats | fit | status | flags.

  3. Score fit — FIRST FILTER (hard gate, nothing filled before it passes). Read resume.* and the JD, give a 0–100 fit score with a 2–3 line rationale (matched strengths, gaps), and record it in the tracker row. Read the user's Minimum fit score to apply from profile.md (default 50 if unset). If the score is below that threshold — or there's a hard-eligibility blocker (the role's country needs work authorization the user lacks) — STOP here: set the tracker row status=skipped (reason: low-fit or ineligible), report the score, the threshold, and why, and do not fill the form. The user can override by replying "apply anyway" or lowering their threshold.

3b. Account / login wall (Workday, iCIMS, SuccessFactors — only after the fit-gate passes; never create an account for a role you won't apply to). If applying requires sign-in / create-account before the form:

  • Already signed in (persistent session) → continue.
  • Saved creds exist → fetch this portal's password from the macOS Keychain (security find-generic-password -s dear-hiring-manager -a "<domain>|<email>" -w) and log in with the profile email + that password.
  • New portal → generate a strong unique password (never reuse across portals), fill create-account with the profile email + that password, submit, then save it to the Keychain (security add-generic-password -U -s dear-hiring-manager -a "<domain>|<email>" -w "<password>"). Passwords live only in the Keychain — never in profile.md / answers.md / any file.
  • Email verification required → set tracker status=blocked, pause and ask the human to click the verify link in their inbox (or, if a job-inbox integration is configured, open + click it), then resume.
  • CAPTCHA / 2FA on signup → do not solve; set blocked, hand to the human, resume after. Then continue to the form.
  1. Map every field — fill or flag, never silently skip. Every required field must end up either filled or explicitly flagged; a blank required field is acceptable only when it is flagged. Paginated wizards (Workday-style: My Information → My Experience → Questions → Self-Identify → Review): run this fill pass on each page, click Next/Continue to advance, re-snapshot after every page, and only stop at the FINAL Submit (step 7) — never click an intermediate Submit. If the portal offers autofill-from-resume, use it first, then complete/fix the parsed fields. For each field:

    • Profile-backed (name, contact, salary): fill from profile.md.
    • Work authorization (per role country): answer for the ROLE's country — do not paste the profile's raw yes/no. E.g. authorized in NL+EU but the role is in the US → "authorized: No", "sponsorship: Yes". Flag whenever the role country is outside the user's authorized regions.
    • Legal attestations (non-compete, felony, illegal activity, accommodation, previously-employed, age): fill only from a confirmed, non-blank profile.md value. If blank/unconfirmed, flag and leave for the human — never auto-attest a legal answer from a default.
    • Voluntary EEO (gender, race/ethnicity, veteran, disability): use profile.md if set; if blank, select "prefer not to answer" / "decline to self-identify". Never flag, never block on these.
    • Open-ended / unseen (e.g. "Why this company?", custom screening): search answers.md for a matching prior Q&A. Prefer the newest, most company/role-specific match.
      • Match found → adapt and fill.
      • No match → write a best-guess grounded in resume + profile, fill it, and flag it. Run every open-ended answer you write or adapt through the humanize skill — sound like the candidate, no AI tells, no em/en-dashes, no parenthetical asides.
    • File uploads (resume/CV, cover letter): Playwright MCP sandboxes file reads to the workspace root. Copy ~/.dear-hiring-manager/resume.* into .playwright-mcp/ (gitignored), upload from there, then delete the staged copy. Do NOT enable --allow-unrestricted-file-access — staging keeps the sandbox intact instead of exposing every file on the machine.
    • Comboboxes: many ATS (Greenhouse/Ashby) render react-select, not native <select>. You MUST first click the combobox to open its menu, THEN type the option text, THEN press Enter. Typing/fill without opening first silently fails — the field stays empty (Select...). Re-snapshot to confirm each value stuck; browser_fill_form with type combobox does not work on react-select.
    • Confidence rule: high confidence → fill quietly. Low confidence or anything legal/sensitive with no profile answer → fill best-guess and flag, or leave blank and flag. When in doubt, flag.
  2. Cover letter (auto-detect + deliver). If the form has a cover-letter field — a textarea OR a file-upload ("Attach" a cover letter) — automatically invoke the cover-letter skill to draft a tailored, truthful letter from resume + JD, then deliver it: fill the textarea, or (for upload) save it to a file and upload it (per the cover-letter skill). Flag it review-me. No such field → skip.

  3. CAPTCHA / anti-bot. If you hit a CAPTCHA or bot check, do not attempt to solve or evade it. Set the tracker row status=blocked (reason: CAPTCHA/bot check), leave the tab as-is, and tell the human to complete that step manually.

  4. Stop before Submit. Leave the browser tab open, parked at the final Submit button (for a multi-page wizard, only after the last page is filled — intermediate Next/Continue is fine). Print a BRIEF review — do NOT dump every filled field. Just:

    • One line: fit N/100 · M fields filled.
    • Flags in RED — one line each, prefixed with 🔴 (terminal markdown has no text color, so 🔴 IS the red): 🔴 <field>: <what you guessed / why> for anything needing the human's eyes, and 🔴 <field>: blank for each required field left empty.
    • If nothing needs attention, say so in one line ("nothing flagged"). Update the tracker row status=filled. Then one line: "Review the 🔴 items, fix, and Submit yourself."
  5. Learn after submit. When the human says they've submitted (or edited), ask what they changed. Route each change to the right place so future applications benefit:

    • Profile-level fact (salary expectation, location/city, phone, links, work authorization — a field that lives in profile.md): update profile.md so every future application picks it up and stops re-flagging it, not just keyword lookups.
    • Company/role-specific or open-ended (e.g. "Why this company?", a custom essay): append a new entry to answers.md (never overwrite), source: human-edit, tagged with company + role + date.
    • Legal attestation the human confirms: write the confirmed value to profile.md (now confirmed) and seed a matching answers.md entry. Then update the tracker row status=submitted.
Show full SKILL.md (181 more words)Show less

Rules

  • AI understands; the browser + deterministic steps execute. Don't improvise clicks beyond filling.
  • Never Submit. Never solve CAPTCHAs. Never fabricate legal/EEO answers.
  • Fill or flag every field; never leave a required field silently blank.
  • Flag internal profile inconsistencies you spot (e.g. phone country code vs stated location).
  • Append-only memory: human edits win by being newer, not by overwriting.
  • One posting per run. Batch is Phase 2.

ATS notes (from dogfooding)

  • Greenhouse (job-boards.greenhouse.io, or iframe-embedded e.g. careers.airbnb.com): fields are in the DOM; screening/EEO are react-select comboboxes (click → type → Enter). Works well. A reCAPTCHA usually guards Submit — fine, the human submits. Verified: Airbnb, Overstory, Transcend.
  • Stale/closed postings: redirect to the board root with ?error=true and no form → stop (step 1).
  • Ashby (jobs.ashbyhq.com): JS single-page app; a static fetch sees only the shell. Drive it with the browser (Playwright renders the JS); don't rely on page-source scraping.
  • Lever (jobs.lever.co): bot-blocks static fetches (HTTP 403). Use the browser.
  • Workday (*.myworkday.com): usually requires creating an account/login before the form, plus custom widgets. Treat as Tier 3 — skip or hand to the human unless already signed in.

© extrasmall0, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/apply of extrasmall0/dear-hiring-manager.

Open the folder on GitHubat commit ff6c26b

Compare with similar skills

Apply next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Apply compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Apply this skillextrasmall0/dear-hiring-manager111—~2.6kAutomated safety check: PassMIT
Backend and Agent Project Selectorlishuangqiang/backend-agent-resume-scout347—~1.4kAutomated safety check: PassApache-2.0
E Taxkazukinagata/shinkoku364—~9.1kAutomated safety check: PassMIT
Jobgptsickn33/agentic-awesome-skills47k2 repos~1.4kAutomated safety check: PassMIT
Technical Job Searchgithub/awesome-copilot40k—~1.2kAutomated safety check: PassMIT
Reactive Resume Builderreactive-resume/reactive-resume44k—~1.5kAutomated safety check: PassMIT

Similar skills

  • Backend and Agent Project Selector

    lishuangqiang/backend-agent-resume-scout

    Finds backend or AI agent projects on GitHub that are worth putting on a resume, checks them against local source and writes a Markdown resume package.

    347 GitHub stars~1.4k tokensUpdated 3 mo ago
    Business, Finance & HRAuto-check passed
  • E Tax

    kazukinagata/shinkoku

    This skill should be used when the user wants to file their tax return electronically via the 確定申告書等作成コーナー (NTA Tax Return Preparation Corner) using Claude in Chrome, Antigravity Browser Sub-Agent…

    364 GitHub stars~9.1k tokensUpdated 28 days ago
    Business, Finance & HRAuto-check passed
  • Jobgpt

    sickn33/agentic-awesome-skills

    Job search automation, auto apply, resume generation, application tracking, salary intelligence, and recruiter outreach using the JobGPT MCP server.

    47k GitHub starsUsed in 2 repos~1.4k tokens
    Business, Finance & HRAuto-check passed
  • Technical Job Search

    github/awesome-copilot

    Official

    A skill your agent uses when a software engineer asks for help with job search tasks: parsing or analyzing a job description, tailoring a CV/resume, writing a cover letter, evaluating a job offer…

    40k GitHub stars~1.2k tokensUpdated today
    Business, Finance & HRAuto-check passed
  • Reactive Resume Builder

    reactive-resume/reactive-resume

    Builds resumes as valid JSON for the open-source Reactive Resume app by interviewing you, and can track job applications through its MCP tools.

    44k GitHub stars~1.5k tokensUpdated yesterday
    Business, Finance & HRAuto-check passed
  • Career-Ops Job Search Center

    career-ops-hq/career-ops

    Routes job-search requests to modes for evaluating offers, scanning portals, generating tailored CVs, tracking applications and drafting outreach, starting from a pasted job URL or description.

    74k GitHub stars~3.3k tokensUpdated yesterday
    Business, Finance & HRAuto-check passed

More from extrasmall0/dear-hiring-manager

  • Apify Collect

    extrasmall0/dear-hiring-manager

    Collect fresh job-posting URLs into ~/.dear-hiring-manager/urls.txt by running an Apify job scraper — the discovery source for /batch.

    111 GitHub stars~1.1k tokensUpdated 2 mo ago
    Auto-check: notes
  • Batch

    extrasmall0/dear-hiring-manager

    Batch job-application filler. An agent skill from extrasmall0/dear-hiring-manager.

    111 GitHub stars~928 tokensUpdated 2 mo ago
    Auto-check passed
  • Board

    extrasmall0/dear-hiring-manager

    Application tracker board. An agent skill from extrasmall0/dear-hiring-manager.

    111 GitHub stars~559 tokensUpdated 2 mo ago
    Auto-check passed
  • Onboarding

    extrasmall0/dear-hiring-manager

    Interview the user and build their job-application profile — identity, contact, work authorization, EEO/demographics, and the standard screening answers (non-compete, felony, sponsorship) — plus…

    111 GitHub stars~1.5k tokensUpdated 2 mo ago
    Auto-check: notes
  • Cover Letter

    extrasmall0/dear-hiring-manager

    Generate a tailored, truthful cover letter from the user's resume, profile, and a specific job description.

    111 GitHub stars~749 tokensUpdated 2 mo ago
    Auto-check passed
  • Humanize

    extrasmall0/dear-hiring-manager

    Make generated prose read like a real person wrote it, not AI.

    111 GitHub stars~552 tokensUpdated 2 mo ago
    Auto-check passed

Works with

Questions about Apply

What does Apply do?

Assisted job-application filler. An agent skill from extrasmall0/dear-hiring-manager. Apply is an agent skill from extrasmall0/dear-hiring-manager. Assisted job-application filler.

When should I use Apply?

Apply fits situations like: whenever the user wants help applying to a specific job; tasks that involve Job search and resumes; tasks that involve Browser testing.

How do I install Apply in Claude Code?

Run `npx skills add extrasmall0/dear-hiring-manager --skill apply -a claude-code`. Or copy the skill folder (skills/apply in extrasmall0/dear-hiring-manager) into .claude/skills/apply in your project. Claude Code loads it when a task matches its description.

How do I install Apply in Codex?

Run `npx skills add extrasmall0/dear-hiring-manager --skill apply -a codex`. Or copy the skill folder (skills/apply in extrasmall0/dear-hiring-manager) into .agents/skills/apply in your project. Codex loads it when a task matches its description.

Can I use Apply in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add extrasmall0/dear-hiring-manager --skill apply -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/apply, .gemini/skills/apply, .github/skills/apply and .opencode/skills/apply in your project.

What does Apply need to run?

SKILL.md names no scripts, command-line tools or credentials: Apply is instructions for the agent only.

Does Apply access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Apply safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Apply use?

Apply is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Apply use?

About 2.6k tokens (SKILL.md is roughly 10k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Apply?

Skills that share tags, products or a category with Apply: Backend and Agent Project Selector (lishuangqiang/backend-agent-resume-scout, 347 stars), E Tax (kazukinagata/shinkoku, 364 stars), Jobgpt (sickn33/agentic-awesome-skills, 47k stars) and Technical Job Search (github/awesome-copilot, 40k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Apply?

extrasmall0 (a GitHub user) maintains it in extrasmall0/dear-hiring-manager, which has 111 GitHub stars. The repository holds 7 skills in this directory. The repository was last updated on July 15, 2026.

Source: extrasmall0/dear-hiring-manager on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.