career-apply
Prepare one job application end to end - research the posting, tailor a resume from the master, draft a cover letter and answers in the user's own writing…
Skill by bendechrai
Chalk Labs
Need help deploying career-apply?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/bendechrai-career-apply
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: documents, writing, research
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/career-apply
What's inside
- Skill name
career-apply- Description
- Prepare one job application end to end - research the posting, tailor a resume from the master, draft a cover letter and answers in the user's own writing voice, and fill the application form. Never submits. Use when the user wants to apply for a specific job, tailor a resume, or write a cover letter for a role.
- License
- MIT
- Instructions
- 79 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# /career-apply - prepare one application, stop before submit Argument: a job posting URL, or a company and role from `Jobs/Job Leads.md`. **Never click submit.** That click is always theirs. Also never answer on their behalf: attestations, consents, work authorisation questions, demographic surveys, and salary figures they have not confirmed stay untouched unless they have explicitly given a standing instruction to fill them. One application per invocation. ## Phase 1 - The posting Pull it from the ATS API rather than trusting a search result or a cached page, since postings get edited and closed without the public page catching up. - Ashby: `api.ashbyhq.com/posting-api/job-board/<org>?includeCompensation=true` - Greenhouse: `boards-api.greenhouse.io/v1/boards/<org>/jobs?content=true` - Lever: `api.lever.co/v0/postings/<org>?mode=json` - SmartRecruiters: `api.smartrecruiters.com/v1/companies/<org>/postings` - Workday: `GET https://<org>.wd<N>.myworkdayjobs.com/wday/cxs/<org>/<site>/job/<job-path>` with `Accept: application/json`, which returns `jobPostingInfo`. The `wd<N>` shard and the `<site>` slug both vary per company and can be read off the careers URL. If none of those apply, open the posting in Chrome and read it there. Confirm it is live. Extract the exact title, compensation band, location and remote terms, the full description, and the list of fields the application form will ask for. Check history: prior applications to this company in `Jobs/`, prior contact in Gmail, and anyone in `People/` who works there. Surface all of it. A previous rejection at the same company matters, because their applicant tracking system remembers even when nobody else does. ## Phase 2 - The note Create `Jobs/<Company>/<role>.md` with `status: preparing`, the saved job specification, and a fit assessment that includes an honest gap list. The gap list is internal. It exists so nothing gets overclaimed, and it is exactly what to review before an interview. ## Phase 3 - Checkpoint with them, before drafting This is the largest quality lever in the whole command, and skipping it to look efficient produces a generic application. Ask: 1. **Do they have a personal connection to this?** Have they used the product, met the team, know someone there, got a story about it. The vault cannot know this and it is frequently the thing that makes an application land. 2. **If they have not used the product**, offer the build-first path: a few days actually using it turns a cover letter from admiration into evidence. Then sequence the application after the build. 3. **Compensation**, for this specific role. Never put a number on a form they have not confirmed. 4. **Level and framing calls** the posting forces: individual contributor or leadership, and how to handle anything awkward in the history. If there is a warm contact at the company, raise contact-first sequencing: message the person before applying. A warm introduction outperforms a strong application, and the conversation often answers blocking questions before the form is even started. ## Phase 4 - Resume 1. Read `Assets/resume-master.md`. Everything on the tailored resume must already exist there. If a fact is missing, add it to the master first, then use it. 2. Select rather than rewrite: pick the bullets whose tags match what this posting actually asks for, and retune them to the posting's own vocabulary without changing what they claim. 3. Every bullet must trace to a `Wins/` file. Anything that cannot be traced does not go on the resume. 4. Check the numbers against each other. Years of experience, team sizes, and metrics must not contradict either the master or the drafted answers. Interviewers do notice, and the contradiction costs more than the claim was worth. 5. Keep to one page unless the field expects otherwise. Verify the rendered output rather than assuming. 6. Feed anything new discovered during drafting back into `resume-master.md`. ## Phase 5 - Cover letter and written answers Read `Voice/Voice Profile.md` first and write to it. This is the entire reason the voice profile exists. - **Lead with the outcome or the judgment, not the mechanics.** Over-explaining implementation reads as less senior, not more. - **Use specific stories.** Pull them from `Wins/`, especially the "why I am proud of it" verbatim. Distribute distinct stories across answers rather than reusing one anecdote three times. - **Do not overclaim.** Watch for paraphrase drift, where "built reference applications for X" quietly becomes "built X". Check every sentence against the win it came from. - **Do not name gaps in the prose.** No "I have not used X but I am eager to learn" hedges. They read as generated and they state the obvious. If a gap is genuinely load-bearing, raise it with the user directly rather than writing it into the letter. Close on a forward-looking note, never on a caveat. - **Strip the tells.** After drafting, do a dedicated pass for meta-narration, "not just X but Y" constructions, "I'm excited to" openings, and everything in the tics-to-avoid section of the voice profile. This pass catches things the drafting pass cannot see. - No em dashes or curly quotes, standard keyboard characters only. - Enforce word limits by counting, not by estimating. Store every draft in the job note. Expect to iterate: their edits win, and you fix only mechanical problems and tell them what changed. ## Phase 6 - Form fill Fill what the vault answers conclusively: name, contact, location, links, most recent employer and title, resume upload, and the approved answer text. Leave for them: account creation, background check and privacy acknowledgements, terms checkboxes, relocation attestations, anything the vault cannot answer conclusively, and the submit button. Watch for forms where a mid-flow button is the real submission. Some applicant tracking systems put the final submit on a sub-form with no review page and no confirmation screen, and no visual cue distinguishing it from the per-page save buttons on earlier steps. When in any doubt about whether a button submits, stop and hand the click over. ## Phase 7 - Handoff Report what is filled, what is left for them as a checklist in the note, and the strongest fit argument in case they get a call tomorrow. Leave the browser tab open. When they confirm they have submitted: set `status: applied`, record the date, update `Jobs/Job Leads.md`, and fold any new facts back into `resume-master.md` and the vault. If they abandon it, keep everything. It seeds the next application of the same type.
From bendechrai/career-brain, MIT.
More from bendechrai/career-brain
- 0.524score
career-ingest
bendechrai
Pull career evidence into the Career vault from Granola meeting transcripts, LinkedIn, Gmail, Google Calendar, GitHub, Drive, Slack, or the Incoming drop…
SkillBFreshness: BClaude Code +9 - 0.524score
career-init
bendechrai
One-time setup for a Career Brain - a local vault of career evidence used for positioning, resumes, and job search.
SkillBFreshness: BClaude Code +9 - 0.524score
career-interview
bendechrai
Interview the user to fill gaps in their career record - achievements, scope, who benefited, what they learned, what they are proud of, and what they do and…
SkillBFreshness: BClaude Code +9 - 0.524score
career-jobs
bendechrai
Find roles that match the user's profile, filter them against their stated preferences and dealbreakers, and explain the fit and the stretch on each.
SkillBFreshness: BClaude Code +9 - 0.524score
career-synthesize
bendechrai
Rebuild the positioning one-pager from the evidence in the Career vault - what the user is demonstrably good at, what people say about them, how they work…
SkillBFreshness: BClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Aug 24, 2026 (2 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/bendechrai/career-brain
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from bendechrai
- 0.524score
career-ingest
bendechrai
Pull career evidence into the Career vault from Granola meeting transcripts, LinkedIn, Gmail, Google Calendar, GitHub, Drive, Slack, or the Incoming drop…
SkillBFreshness: BClaude Code +9 - 0.524score
career-init
bendechrai
One-time setup for a Career Brain - a local vault of career evidence used for positioning, resumes, and job search.
SkillBFreshness: BClaude Code +9 - 0.524score
career-interview
bendechrai
Interview the user to fill gaps in their career record - achievements, scope, who benefited, what they learned, what they are proud of, and what they do and…
SkillBFreshness: BClaude Code +9 - 0.524score
career-jobs
bendechrai
Find roles that match the user's profile, filter them against their stated preferences and dealbreakers, and explain the fit and the stretch on each.
SkillBFreshness: BClaude Code +9 - 0.524score
career-synthesize
bendechrai
Rebuild the positioning one-pager from the evidence in the Career vault - what the user is demonstrably good at, what people say about them, how they work…
SkillBFreshness: BClaude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get career-apply running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
career-apply skill. Agent Mart. https://agentmart.dev/skills/bendechrai-career-apply (last updated 2026-10-24)