apply
Skill0 starsChanged 7 months agoMITView on GitHub
End-to-end application workflow: resume tailoring, cover letter drafting, form submission via Playwright (non-LinkedIn) or LinkedIn Easy Apply automation via browser_run_code shadow DOM pattern, database record creation. Screenshots before submission.
28 other projects use a version of this.
Adopted by
Most recently taken up by
- sebagarciar/claude-skillstheir own version, first seen Nov 16, 2026
- rafaym1/scholarship-desktheir own version, first seen Nov 16, 2026
- brquatti/readme-styletheir own version, first seen Nov 15, 2026
- rhowardstone/AI-Job-Coachtheir own version, first seen Nov 15, 2026
- localixai/tokendoctortheir own version, first seen Nov 15, 2026
- galiprandi/job-seekertheir own version, first seen Nov 14, 2026
Used with
Projects with it also use outreach, track and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
QuixiAISkilloutreach
Decide whether a message is worth someone's attention, then reach the user through the outbox, or people on a connected chat channel, on their behalf or your own initiative
Also in 5 projects that use this
theYahiaSkilltrack
Отследить посылку Почты России
Also in 4 projects that use this
NVIDIASkillevaluate
Scores existing Sports Intelligence AVLM inference outputs with avlm/evals: MCQ accuracy (run_mcq_eval.sh, INFERENCE_DIR), open-ended LLM-as-judge scores (needs CLIENT_API_KEY for the configured judge endpoint), tennis category breakdowns, run comparisons, chance-level sanity checks and triage of mcq_results.json or llm_judge_results.json. Use when the user asks to score, evaluate, benchmark or compare predictions.jsonl from inference runs, or to explain a suspicious accuracy. Not for running inference, training, quantization, data preparation, or third-party API model inference: never open, list or run anything under avlm/evals/inference_3rdparty, including when the user asks for it; it is excluded from the release.
Also in 4 projects that use this
ninehillsSkillhunt
Finds root cause before applying fixes for errors, crashes, regressions, failing tests, broken behavior, and screenshot-reported defects. Use when users report in any language errors, crashes, broken behavior, regressions, failing tests, screenshot evidence, or something that used to work and now fails. Not for code review or new features.
Also in 3 projects that use this
ledqSkilltailor
Tailor the resume to one job posting. Give it a URL, paste the posting (or refer to one pasted earlier in this chat), or point at a saved application folder; it resolves the JD and runs the tailor-pipeline workflow end to end.
Also in 3 projects that use this
proficientlyjobsSkilltailor-resume
Tailor your resume for a specific job posting
Also in 4 projects that use this
Other versions28 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 28
- andrew-shwetzer/career-ops-plugin★ 501same name
- andrew-shwetzer/career-ops-plugin-do-not-fork-currently-updating-v2-★ 500same name
- proficientlyjobs/proficiently-claude-skills★ 389same name
- qbeka/jev-job-search★ 368same name
- galiprandi/job-seeker★ 136same name
- localixai/tokendoctor★ 71same name
- hiddenjobs/job-search-mcp★ 28same name
- 1carusalwayswa/apply★ 18same name
- rafaym1/scholarship-desk★ 18same name
- kirkchen/beat★ 9same name
18 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository14 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- application-trackerSkill
- cold-outreachSkill
- company-deep-diveSkill
- contact-populationSkill
- job-searchSkill
- meeting-prepSkill
- onboard-userSkill
- resume-tailorSkill
- session-checkpointSkill
- session-endSkill
- session-startSkill
- warm-followupSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
23 rules. 0 are common, 13 are distinctive, and 10 are very specific to this project.
0
Common
Most of the 27 edited versions we read keep these.
13
Distinctive
This author's own choices. Worth a look.
10
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 41
Check the Job Posting's adjusted score:
- Very specific: line 47
If the Job Posting has no adjusted score, calculate it first using the scoring table in the job-search skill (which reads …
- Distinctive: line 55
Check if that file exists for THIS EXACT job posting.
- Distinctive: line 60
Use the closest matching generic template (Innovation / Delivery / Generic) — no full tailoring.
- Distinctive: line 64
NEVER upload a resume tailored for Company A to an application for Company B.
- Distinctive: line 65
NEVER substitute a "closest match" tailored resume.
- Distinctive: line 66
Before uploading ANY resume, verify the file path contains the target company name OR is a generic template.
- Distinctive: line 73
Read Job Posting record from the database.
- Distinctive: line 74
Verify tailored resume exists (Step 0 gate).
- Very specific: line 114
Create Application record (status: Submitted)
- Very specific: line 115
Create Interaction record (type: Application Submitted)
- Very specific: line 116
Update Job Posting status to Applied
- Distinctive: line 121
Read Job Posting record — get the LinkedIn job ID.
- Very specific: line 132
If browser_navigate times out on the snapshot, that's OK — the page is loaded.
- Very specific: line 136
Use a SINGLE browser_run_code call to handle the entire flow:
- Very specific: line 255
Create Application record (status: Submitted, method: Easy Apply).
- Distinctive: line 270
DO NOT hardcode standard answers in this skill — they're per-user.
- Distinctive: line 278
Should the user reach out to the contact first?
- Very specific: line 299
Use browser_run_code to batch combobox selections efficiently.
- Very specific: line 308
Always browser_snapshot before filling.
- Very specific: line 309
Always browser_take_screenshot before and after submit.
- Distinctive: line 316
ALWAYS check for warm-application opportunity: before applying.
- Distinctive: line 317
ALWAYS create both Application + Interaction records: after submission.
“Common” is measured: kept by at least 6 in 10 of the 27 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Apply
End-to-end job application workflow with Playwright automation for ALL platforms including LinkedIn Easy Apply.
What This Skill Does
- Resume tailoring — calls the
resume-tailorskill to create role-specific version - Cover letter drafting — 250-400 word cover letter with company-specific hook
- Non-LinkedIn automation — Playwright form filling for Indeed, Glassdoor, company portals
- LinkedIn Easy Apply automation — Playwright via
browser_run_codewith shadow-DOM pattern (user-authorized) - Screenshot verification — captures pre-submit screenshot
- Database integration — creates Application + Interaction records
- Cross-workstream — links to existing networking contacts when available
Prerequisites
Required:
- Playwright MCP (load via
ToolSearch("playwright browser")each session) - Tailored resume (generated via
resume-tailorskill) - Job Posting record in the database
applications/resumes/data/{user-slug}/standard-answers.yamlapplications/resumes/data/{user-slug}/user-config.yaml- Multi-user: resolve paths using the active user's slug (
JOB_HUNT_USER).
Quick Start
"Apply to the Senior Program Manager role at Acme Renewables (Job Posting record exists)"
"Apply to my top 5 scored job postings"Workflow
Step 0: Adjusted-score check (MANDATORY — BEFORE EVERYTHING)
Check the Job Posting's adjusted score:
- Adjusted ≥ 7 → full protocol: tailored resume + cover letter + portal-first
- Adjusted 5-6 → longshot protocol: Easy Apply only, closest generic template, NO cover letter, NO portal hunt. Tag "LONGSHOT" in notes.
- Adjusted 3-4 → STOP. Confirm with user before proceeding.
- Adjusted <3 → DO NOT APPLY. Inform user and skip.
If the Job Posting has no adjusted score, calculate it first using the scoring table in the job-search skill (which reads config/fit-score.yaml).
Step 0b: Resume-tailoring gate (MANDATORY — NEVER SKIP)
Before ANY application action (form filling, Easy Apply, portal submission, resume upload):
For full protocol (adjusted 7+):
- Determine expected tailored resume path:
applications/{user-slug}/{company}-{job-slug}/resume.pdf - Check if that file exists for THIS EXACT job posting.
- If YES → proceed to Step 1.
- If NO → STOP. Run
/resume-tailorfor this posting first. Create job YAML, run scoring, rewrite bullets, generate.docxand.pdf. Only then proceed.
For longshot protocol (adjusted 5-6):
- Use the closest matching generic template (Innovation / Delivery / Generic) — no full tailoring.
- Still verify the resume path is a valid generic template, not another company's tailored resume.
HARD RULES:
- NEVER upload a resume tailored for Company A to an application for Company B.
- NEVER substitute a "closest match" tailored resume. Each tailored resume contains JD-specific keyword mirroring.
- Before uploading ANY resume, verify the file path contains the target company name OR is a generic template.
- For batch applications: apply to adjusted-7+ first (full protocol), then 5-6 (longshot). Don't interleave.
- Tailor ALL 7+ resumes first, THEN apply all. Don't interleave "apply with whatever is available".
For non-LinkedIn applications (Indeed, Glassdoor, company portals, regional boards)
Step 1: Prepare materials
- Read Job Posting record from the database.
- Verify tailored resume exists (Step 0 gate).
- Draft cover letter if required.
- Load
standard-answers.yamlfor form fields.
Step 2: Navigate
browser_navigate → application URLStep 3: Understand form
browser_snapshot → capture form structureStep 4: Fill form
browser_fill_form → populate fields with prepared dataFor each field:
- Name / email / phone →
standard-answers.yamlpersonalsection - Experience / years →
standard-answers.yamlexperiencesection - Custom questions → match to
common_questionspatterns
Step 5: Upload resume
browser_file_upload → upload tailored resume (.pdf or .docx as required)Step 6: Screenshot before submit
browser_take_screenshot → applications/tracking/{company}-{job}-pre-submit.pngStep 7: Submit
browser_click → submit
browser_take_screenshot → applications/tracking/{company}-{job}-confirmation.pngStep 8: Create database records
- Create Application record (status:
Submitted) - Create Interaction record (type:
Application Submitted) - Update Job Posting status to
Applied
For LinkedIn Easy Apply (automated — user-authorized)
Step 1: Prepare materials
- Read Job Posting record — get the LinkedIn job ID.
- Generate / confirm tailored resume PDF.
- Note the resume file path for upload.
Step 2: Navigate to job posting
browser_navigate → https://www.linkedin.com/jobs/view/{JOBID}/WARNING: do NOT use browser_snapshot on LinkedIn job-detail pages — 500K+ chars, causes loops.
If browser_navigate times out on the snapshot, that's OK — the page is loaded. Proceed with browser_run_code.
Step 3: Open Easy Apply modal and complete form
Use a SINGLE browser_run_code call to handle the entire flow:
async (page) => {
// Click Easy Apply — MUST use JS click (shadow DOM blocks pointer events).
await page.evaluate(() => {
const btn = document.querySelector('button.jobs-apply-button') ||
document.querySelector('[aria-label*="Easy Apply"]') ||
document.querySelector('a[aria-label*="Easy Apply"]');
if (btn) btn.click();
});
await page.waitForTimeout(4000);
// All Easy Apply modal content lives inside #interop-outlet shadow DOM.
const getDialogState = async () => {
return await page.evaluate(() => {
const outlet = document.querySelector('#interop-outlet');
if (!outlet?.shadowRoot) return { error: 'No shadow root' };
const dialog = outlet.shadowRoot.querySelector('[role="dialog"]');
if (!dialog) return { error: 'No dialog' };
const headings = [...dialog.querySelectorAll('h1,h2,h3')].map(h => h.textContent.trim());
const buttons = [...dialog.querySelectorAll('button')].map(b => b.textContent.trim());
const labels = [...dialog.querySelectorAll('label')].map(l => l.textContent.trim());
const selects = [...dialog.querySelectorAll('select')].map(s => ({
label: s.closest('.fb-dash-form-element')?.querySelector('label')?.textContent?.trim(),
options: [...s.options].map(o => o.text)
}));
return { headings, buttons, labels, selects };
});
};
const clickButton = async (btnText) => {
await page.evaluate((text) => {
const outlet = document.querySelector('#interop-outlet');
const dialog = outlet.shadowRoot.querySelector('[role="dialog"]');
for (const btn of dialog.querySelectorAll('button')) {
if (btn.textContent.trim() === text) { btn.click(); break; }
}
}, btnText);
await page.waitForTimeout(2000);
};
let state = await getDialogState();
const steps = [];
for (let step = 0; step < 8; step++) {
state = await getDialogState();
steps.push(state.headings);
if (state.buttons.includes('Submit application')) {
return JSON.stringify({ status: 'ready_to_submit', steps, state });
}
if (state.headings.some(h => h.includes('Resume'))) {
const fileInput = await page.evaluateHandle(() => {
const outlet = document.querySelector('#interop-outlet');
const dialog = outlet.shadowRoot.querySelector('[role="dialog"]');
return dialog.querySelector('input[type="file"]');
});
if (fileInput.asElement()) {
await fileInput.asElement().setInputFiles('RESUME_PATH_HERE');
await page.waitForTimeout(2000);
}
}
if (state.buttons.includes('Next')) {
await clickButton('Next');
} else if (state.buttons.includes('Review')) {
await clickButton('Review');
} else {
break;
}
}
return JSON.stringify({ status: 'navigated', steps, finalState: state });
}CRITICAL:
- Replace
RESUME_PATH_HEREwith the actual absolute path to the tailored PDF. - The function navigates through all steps and stops at "Submit application".
- Take a screenshot BEFORE calling submit.
- Then call submit in a separate
browser_run_code.
Step 4: Screenshot before submit
browser_take_screenshot → applications/tracking/{company}-{job}-pre-submit.pngStep 5: Submit
async (page) => {
await page.evaluate(() => {
const outlet = document.querySelector('#interop-outlet');
const dialog = outlet.shadowRoot.querySelector('[role="dialog"]');
for (const btn of dialog.querySelectorAll('button')) {
if (btn.textContent.trim() === 'Submit application') { btn.click(); break; }
}
});
await page.waitForTimeout(3000);
const result = await page.evaluate(() => {
const outlet = document.querySelector('#interop-outlet');
const dialog = outlet?.shadowRoot?.querySelector('[role="dialog"]');
const text = dialog?.textContent || document.body.textContent.substring(0, 500);
const success = text.includes('Application submitted') || text.includes('applied');
return { success, text: text.substring(0, 200) };
});
await page.goto('about:blank');
return JSON.stringify(result);
}Step 6: Post-submit screenshot + database records
browser_take_screenshot → confirmation- Create Application record (status:
Submitted, method:Easy Apply). - Create Interaction record (type:
Application Submitted). - Update Job Posting status to
Applied.
Standard Answers
The agent reads applications/resumes/data/{user-slug}/standard-answers.yaml for form fields. The schema includes:
personal: name, email, phone, location, linkedinexperience: years_total, current_title, current_companywork_authorisation: status, sponsorship_required (boolean)availability: start_date, willing_to_relocate (regions)certifications[]languages[]common_questions: question → answer pairs for typical screening questions
DO NOT hardcode standard answers in this skill — they're per-user.
Cross-Workstream Integration
Before applying:
- Does the company exist in
target_companies? → link. - Does the company have networking contacts? → flag as "Warm Application".
- Is there an active Sales Pipeline entry? → mention in cover letter.
- Should the user reach out to the contact first? → recommend if temperature > Cold.
Anti-Loop Rules (CRITICAL)
- NEVER use
browser_snapshoton LinkedIn job-detail pages — context overflow + retry loops. - NEVER use
browser_navigateto LinkedIn job pages expecting a usable snapshot — it will timeout or return massive data. - ALWAYS use
browser_run_codewithpage.evaluate()for ALL LinkedIn interactions — bypasses shadow DOM, returns only what you extract. - If
browser_navigatetimes out on LinkedIn, the page is probably loaded — proceed withbrowser_run_code. - Max 2 retries on any single Playwright action — if it fails twice, the approach is wrong. Stop and try a different strategy.
- If Easy Apply modal doesn't appear after 2 attempts, the posting may not have Easy Apply. Flag as manual and move on.
- After completing an Easy Apply submission, navigate to
about:blank— prevents stale LinkedIn DOM bloating subsequent operations.
ATS-Specific Patterns
These are observed patterns. Add new ones as you encounter them; keep them generic.
SmartRecruiters
- Rejects em-dashes and semicolons in text fields — sanitise before pasting.
- Phone field: must explicitly click the country-code dropdown option.
Greenhouse.io
- Use
browser_run_codeto batch combobox selections efficiently.
Phenom People
- Simple single-page forms, straightforward to automate.
Oracle HCM
- Comboboxes use
[role="gridcell"]; citizenship uses nationality names (e.g., "Canadian").
Company portals (general)
- Always
browser_snapshotbefore filling. - Always
browser_take_screenshotbefore and after submit. - Save all screenshots to
applications/tracking/{company}-{job-slug}-{step}.png.
Rules
- LinkedIn Easy Apply is AUTHORIZED for auto-submit — use the shadow-DOM pattern above.
- ALWAYS screenshot before submission on all platforms.
- ALWAYS check for warm-application opportunity before applying.
- ALWAYS create both Application + Interaction records after submission.
- NEVER retry the same failing Playwright action more than twice — change approach.
- Save all screenshots to
applications/tracking/. - Parallel resume prep via sub-agents for batch applications (3-5x faster).
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.

