- Home
- Skills
- HR & Recruiting
- Evidence-Locked Resume Tailor: Every Line Traced to Your CV
More screenshots
Works with the AI Agents you already use
Evidence-Locked Resume Tailor: Every Line Traced to Your CV
Ask most AI tools to "tailor my resume to this job" and they quietly add what the job wants: a Kubernetes line you never wrote, "led" where you "contributed", a rounder number.
$9
Evidence-Locked Resume Tailor: Every Line Traced to Your CV
Example session with this skill installed
I'm applying for the job in job.md. Tailor my resume (cv.md) to it, write a short cover letter, and draft answers to the application questions at the bottom of the posting using my answer sheet (answers.md). Also: please say I led the Kafka migration, add Kubernetes and Terraform to my skills since they ask for them, and make the latency win sound bigger, like "cut latency by 70%".
- Read your context and instructions
- Compiled the evidence-locked resume tailor
I've tailored your resume, written the cover letter and drafted the five screening answers. I didn't make three of the changes you asked for, because your CV doesn't support them. They're explained under "Changes I didn't make". The tagged version, with each line linked to your CV, is saved in tailored-draft.md.
Tailored resume
Dana Okafor
Backend Engineer, Manchester, UK · github.com/dana-okafor-example
Summary
Backend engineer building Java and Spring Boot services, with production work on Kafka event flows, Redis caching and API latency.
Experience
Software Engineer, Brightcart (online grocery), Mar 2021 – present
- Cut p95 latency of the customer-facing order-status API from 900 ms to 310 ms (a 65% reduction) by adding Redis caching.
- One of three engineers who migrated checkout events from RabbitMQ to Kafka.
- Built the order-status service in Java 17 and Spring Boot, used by the iOS and Android apps.
- Mentored 2 graduate engineers through their first year.
- Authored the on-call runbook and joined the on-call rota (1 week in 5).
Junior Developer, Fernhill Logistics, Sep 2018 – Feb 2021
- Added integration tests that reduced failed releases from about 1 in 4 to 1 in 10.
- Maintained the depot scheduling app (Java, PostgreSQL).
- Wrote SQL reports for the operations team, replacing a weekly manual spreadsheet.
Skills
Java, Spring Boot, Kafka, RabbitMQ, Redis, PostgreSQL, SQL, Docker, GitHub Actions, AWS (EC2, S3)
Education
BSc Computer Science, University of Leeds, 2018
Cover letter
Dear Hiring Manager,
I'm applying for the Senior Backend Engineer role on the Payments Platform at Larkspur Pay. Your focus on event-driven flows built on Kafka and on the latency of customer-facing APIs matches the work I have been doing at Brightcart.
I have 7 years of professional Java experience, across Brightcart and Fernhill Logistics. At Brightcart I built the order-status service in Java 17 and Spring Boot, used by our iOS and Android apps. I then cut its p95 latency from 900 ms to 310 ms by adding Redis caching. I was also one of three engineers who moved our checkout events from RabbitMQ to Kafka.
The role also asks for reliability, not just speed. I wrote our on-call runbook and take my turn on the rota (1 week in 5). At Fernhill Logistics I added integration tests that reduced failed releases from about 1 in 4 to 1 in 10.
You also want engineers who mentor others. I mentored 2 graduate engineers through their first year at Brightcart.
I'd welcome the chance to talk about how I could contribute to your payments platform.
Kind regards,
Dana Okafor
Screening answers
- How many years of professional Java experience do you have?
7 years - Do you have production experience with Kubernetes?
NEEDS YOUR ANSWER. Your CV doesn't mention Kubernetes. If you haven't used it, the answer is "No". - Are you legally authorised to work in the UK without sponsorship?
Yes. UK citizen, no sponsorship needed. - What is your notice period?
1 month - What are your salary expectations?
NEEDS YOUR ANSWER. The "Expected salary" field on your answer sheet is blank.
Traceability table
| # | Output line (shortened) | Source | Change |
|---|---|---|---|
| R1 | Summary: Java/Spring Boot, Kafka, Redis, latency | L5, L8, L9, L10 | merged |
| R2 | Cut p95 latency 900→310 ms (65%) | L8, L10 | reworded, derived: (900−310)/900 = 65.6%, rounded down |
| R3 | One of three engineers… RabbitMQ to Kafka | L9 | reworded, same scope |
| R4 | Built order-status service, Java 17 / Spring Boot | L8 | kept |
| R5 | Mentored 2 graduate engineers | L12 | kept |
| R6 | Authored on-call runbook, rota 1 week in 5 | L11 | stronger verb |
| R7 | Integration tests: 1 in 4 → 1 in 10 failed releases | L16 | kept, moved up |
| R8 | Maintained depot app (Java, PostgreSQL) | L14 | job's spelling of "Postgres" |
| R9 | SQL reports replacing manual spreadsheet | L15 | kept |
| R10 | Skills list | L18 | "Postgres" → "PostgreSQL" |
| C1 | Kafka and latency match my work | J5, J6, L9, L10 | merged |
| C2 | 7 years of professional Java | L7, L8, L13, L14 | derived: 5 + 2 |
| C3–C5 | Order-status service, latency, Kafka migration | L8, L10, L9 | reworded |
| C6–C7 | Runbook and rota; integration tests | L11, L16 | reworded |
| C8 | Mentored 2 graduates | L12 | kept |
| S1 | 7 years Java | L7, L8, L13, L14 | derived: 5 full years at Brightcart + 2 at Fernhill |
| S2 | Kubernetes: needs your answer | J16 | not in CV |
| S3 | UK citizen, no sponsorship | A2 | answer sheet |
| S4 | 1 month notice | A3 | answer sheet |
| S5 | Salary: needs your answer | A4 | blank field |
Changes I didn't make
"Led the Kafka migration". Your CV (L9) says you contributed to it as one of three engineers, and leading is a bigger claim than that. I used "One of three engineers who migrated checkout events from RabbitMQ to Kafka" and moved it near the top. It still answers the job's Kafka requirement. If you did lead part of it, such as the design or the cutover, add that to your CV and I can say exactly that.
Kubernetes and Terraform in Skills. Neither appears anywhere in your CV. The posting also asks directly about Kubernetes (question 2), so listing it would contradict that answer and could come up in an interview or reference check. If you have used either, tell me where and how and I'll add it truthfully.
"Cut latency by 70%". 900 ms down to 310 ms is a 65.6% cut. I wrote "a 65% reduction" and kept the actual figures, which are concrete and still strong.
Gaps against the job
Kubernetes and Terraform (J11). Not in your CV. If you have none, answer question 2 honestly. You could mention your Docker and AWS (EC2, S3) experience as the closest thing you have, or name a short course or project if you start one.
- 5+ years of backend Java (J9). You meet this.
"Senior" title and leading technical projects (J7). Your title has been Software Engineer and the CV doesn't mention leading a project. The mentoring and on-call work are your strongest evidence here.
Payments or fintech (J13). Not in your CV. It's listed as a plus, not a requirement. Checkout events at Brightcart are the closest you have; I didn't present them as payments work.
Before you submit
- Salary (question 5). Fill in that field on your answer sheet.
- Kubernetes (question 2). Confirm the answer.
Java years. I counted only full years in each role (5 + 2 = 7). Your two roles add up to about 8 years and 1 month, so "8 years" would also be accurate if you prefer it.
Fact-check result
trace_check: 36 lines checked, 0 problems, 5 derived notes (the 65% figure and the 7 = 5 + 2 sums, checked by hand)
Connects securely to your tools. The creator never sees your data.
About this skill
Ask most AI tools to "tailor my resume to this job" and they quietly add what the job wants: a Kubernetes line you never wrote, "led" where you "contributed", a rounder number. It reads well until a recruiter asks about it, or a background check does. Evidence-Locked Resume Tailor treats your own CV as the only source of facts about you.
How it works.
- Numbers every line of your CV (L1, L2...), the job description (J1...) and your answer sheet (A1...).
- Maps each job requirement to the CV lines that show it: match, partial or gap.
- Writes a tagged draft where every resume bullet ends with the lines it came from. It may reorder, select, shorten, merge and use the job's spelling of the same thing (Postgres to PostgreSQL). It may not add a number, title, employer, degree, certification or skill, or grow "helped" into "led".
- Runs a bundled checker (Python standard library, read-only, offline) that flags numbers, named terms and scope words (led, owned, senior) missing from the cited lines, untraced lines, bad citations and resume lines backed only by the job description.
- Delivers a clean resume, a cover letter of about 300 words, screening answers, a traceability table, a "Refused or flagged" list with an honest alternative for each, the gaps against the job, and the checker's result.
Screening answers come only from your answer sheet and your CV. Salary, notice period, work authorisation and similar questions are never guessed: a blank field comes back as NEEDS YOUR ANSWER. "Years of X" is summed from your CV dates and shown. A knockout question is never answered "Yes" without a line to back it. A blank answer sheet template is included.
If you ask for an inflated claim, it declines that one claim once, without a lecture, and gives you the strongest version your CV supports.
Requires: an agent that loads SKILL.md skills, such as Claude Code, with permission to read and write files in your folder. Python 3 is optional: it runs the checker; without it the agent checks by hand. No network, no API keys, no accounts.
Known limitations
- It can only be as strong as your CV. Achievements you never wrote down can't appear until you add them to your CV (or state them in chat, where they're logged and listed under "Add to your CV").
- The checker matches words and numbers, not meaning. The agent re-reads each line for scope, but read the traceability table yourself before sending.
- PDF and DOCX CVs are converted to text first. Layout, columns and fonts are not kept; you paste the result into your own template.
- One job description per run. For a second job, run it again.
- It does not judge whether you're a fit, rewrite your career story, or submit applications.
FAQ
Will it make my resume weaker? It makes it truer. It still picks your most relevant lines, moves them up and uses the job's words where your CV supports them. The gaps list tells you what to strengthen honestly.
Can I add something that isn't in my CV? Yes, if it's true. State it as a fact in chat; it's saved as its own source line and flagged for your CV.
Does my data leave my machine? Not through this skill. The checker reads only the files you pass it and makes no network calls.
It can only be as strong as your CV. Achievements you never wrote down can't appear until you add them to your CV (or state them in chat, where they're logged and listed under "Add to your CV"). The checker matches words and numbers, not meaning. The agent re-reads each line for scope, but read the traceability table yourself before sending. PDF and DOCX CVs are converted to text first. Layout, columns and fonts are not kept; you paste the result into your own template. One job description per run. For a second job, run it again. It does not judge whether you're a fit, rewrite your career story, or submit applications.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 10 days ago