cv-tailor
ACTIVATE when producing or reworking a CV, a résumé or an ESN dossier de compétences for the French market, or when adapting one to a specific job posting or mission brief. ACTIVATE for 'refais mon CV', 'adapte mon CV à cette offre', 'dossier de compétences', 'mon CV fait trop de pages', 'mon CV ne passe pas les ATS', 'CV freelance', 'tailor my resume', 'ATS check'. Compiles the career-evidence referential into a two-page ATS-safe CV or a mission-by-mission dossier, applies parsing rules taken from vendor documentation rather than folklore, and can never emit a client name whose disclosure level forbids it. DO NOT use to build or update the underlying referential (see career:career-evidence), to rewrite a LinkedIn profile (see career:linkedin-profile), or to write a cover letter or a commercial proposal.
- License 许可
- MIT
- 最近更新
- 2026-10-13
- Stars
- ★ 9
- Forks
- 2
01 前言 Front Matter
- name
- cv-tailor
- description
- ACTIVATE when producing or reworking a CV, a résumé or an ESN dossier de compétences for the French market, or when adapting one to a specific job posting or mission brief. ACTIVATE for 'refais mon CV', 'adapte mon CV à cette offre', 'dossier de compétences', 'mon CV fait trop de pages', 'mon CV ne passe pas les ATS', 'CV freelance', 'tailor my resume', 'ATS check'. Compiles the career-evidence referential into a two-page ATS-safe CV or a mission-by-mission dossier, applies parsing rules taken from vendor documentation rather than folklore, and can never emit a client name whose disclosure level forbids it. DO NOT use to build or update the underlying referential (see career:career-evidence), to rewrite a LinkedIn profile (see career:linkedin-profile), or to write a cover letter or a commercial proposal.
02 资源文件 4
- 参考文档index.json512 B
- 参考文档references/ats-rules-fr.md8.9 KB
- 参考文档references/bullet-formulas.md10.1 KB
- 参考文档references/dossier-competences.md6.1 KB
03 SKILL.md 正文
CV Tailor
Compile the career-evidence referential into a document aimed at one reader. The
referential is the source and is never edited from here: a fact missing at compile
time is a fact to capture upstream, not to invent downstream.
1. Classify before writing
Never produce a document before these four are settled. They decide length, structure, vocabulary and what may be named. Ask only for what cannot be read from the request.
| Values | Decides | |
|---|---|---|
| Reader | intermediary (recruiter, ESN sales, RH) · technical decision-maker (CTO, tech lead) · training buyer | vocabulary, what leads |
| Artifact | 2-page CV · dossier de compétences · per-posting variant | structure and length |
| Channel | ATS form · direct email · freelance platform · ESN | parsing constraints, anonymisation |
| Target | the role aimed at, in the market's own words | title, ordering, keyword set |
The reader is the axis most often collapsed. On the French market a large share of inbound goes through intermediaries who filter on vocabulary without reading the technology. The same evidence therefore compiles differently: a technical decision-maker wants the problem and the architecture, an intermediary wants the role title, the stack and the availability.
2. Two artifacts, one source
These are different documents, not two lengths of the same one. Producing a compromise between them is the most common failure and it serves neither reader.
The CV — two pages, never three. Beyond ten years of experience one page under-delivers, and past two pages a measurable share of recruiters reject on sight. Detail the last five to seven years; condense everything before into a single dense block. Structure: identity → headline and summary → key skills → selected engagements → earlier career → training and certifications.
The dossier de compétences — long, mission by mission. The format French intermediaries expect, and the one they can re-skin. Roughly half the volume goes to engagements, each opening with client context before its outcomes. Detail on the last five to seven years, micro-engagements grouped rather than listed one by one. See references/dossier-competences.md.
Cutting an eleven-page dossier down to two pages destroys the artifact intermediaries actually read. Compile both from the same referential instead.
3. Aggregate what is repetitive, detail what is singular
Engagements are the reason someone reads the document, so they carry the detail. Recurring activity is proved by its totals, not by its list: twenty-five training sessions become four lines, seven audits become three. The referential keeps every entry so the totals stay recomputable.
Aggregate training by course rather than by year — a reader wants to know what can be taught, and a chronological cut exposes gaps that need a sentence rather than a table. Prefer participant counts over session counts when a course was delivered to large cohorts, and state the totals so nothing looks concealed.
Where a course or an offering is no longer sold, keep it in a separate line rather than folding it into the headline figure.
4. Disclosure is enforced, not advised
Before emitting any client name, read its disclosure in the referential.
public→ the name may appear.sector-only→ emit## Formulation publiqueverbatim. Never the name, never the product, never the acronym, never a domain narrow enough to identify the buyer.excluded→ the engagement contributes to totals, seniority and date continuity, and to nothing else.
If a restricted entry has no ## Formulation publique, stop and say so. Do not
sanitise on the fly: judging what identifies a buyer in a narrow market is the person's
call, made once, upstream.
5. Bullets
One formula, applied everywhere: verb, then object with the technology named, then a figure or a scope, on the same line. The reader scans line starts, so the verb comes first and the figure never spills to the next line.
Where no figure exists, state the scope instead of inventing an outcome, and make the challenge explicit so the bullet still carries weight. Two lines is the right density; one is thin, three is the ceiling.
Every technology listed in the skills section must appear in at least one engagement bullet. A skill listed and never demonstrated is what reviewers call inflated duties, and it is one of the markers they read as machine-written. See references/bullet-formulas.md.
6. Keyword coverage is a floor
Extract the posting's vocabulary, classify each term as covered, partial, gap or not-applicable, and place what is missing where the corresponding filter reads it. Target roughly three quarters coverage and stop there.
Full coverage is a defect, not a goal. It means terms were claimed without evidence, and reviewers name keyword saturation as the first thing that makes a document read as generated. Write the acronym and its expansion once each. Never append a keyword block at the end of the document: it ties no term to a dated engagement and reads as stuffing.
On form-based channels, fill the skills and languages fields by hand. At least one major ATS never populates them from the uploaded file, so the filters a recruiter uses stay empty however good the document is.
7. Gate before delivery
Run these in order. The first three are mechanical and must pass.
- Plain-text extraction. Extract the text from the produced file and read it. That
is what a parser sees. Wrong order, merged columns or words that lost letters mean
the file is broken whatever it looks like on screen. Watch the
fiffflligatures in French: certification, profil, efficacité, qualifications. - Parsing rules. Single column, no tables, no text boxes, nothing in headers or footers, contact details in the body, standard section headings, one date format, text-layer file under 2.5 MB. Full list with sources: references/ats-rules-fr.md.
- Disclosure. No restricted name anywhere in the output.
- Dates. No unexplained overlap, no unexplained gap, and consistency with the LinkedIn profile — recruiters check both.
- Human pass. Read it aloud. Every verb should be one the person would actually say. Vary bullet length deliberately; uniform rhythm is a tell.
Do not compute or report an ATS score. No major vendor produces one a candidate could aim at — scores come from the tools that sell them. Report findings, not a grade.
8. Deliver with its diff
Ship the document plus what changed and why, section by section, one line per change. The person accepts or rejects each one. A document rewritten in one block cannot be reviewed, only trusted.
State explicitly what was left out and why: entries dropped for relevance, figures missing upstream, sections aggregated. Silent truncation reads as full coverage.
Anti-patterns
- Compromising between the CV and the dossier. Two readers, two artifacts.
- Reporting a score. It measures the tool that computed it.
- Sanitising a restricted client on the fly instead of stopping when the public formulation is missing.
- Maximising keyword coverage. Past the floor it is the marker reviewers reject on.
- Editing the referential from here. Missing facts go upstream.
- Reordering engagements out of chronology to put a relevant one first. Parsers and readers both infer recency from position; move the bullets, not the entries.
- Reducing legibility to hit a page count. A clean two pages beats a cramped one.
04 同仓库更多 Skills
vitest-test-conventions
ACTIVATE when writing Vitest tests, creating test files, using vi.fn()/vi.mock()/vi.spyOn(), or test factories. ACTIVATE for 'Vitest', 'it.each', 'NestJS testing module', 'supertest', 'vi.fn', 'vi.mock', 'spyOn'. Provides Vitest/NestJS-specific testing patterns; cross-language testing principles (DAMP, AAA, spy over mock, what NOT to test) live in craft:testing-principles. DO NOT use for: node:test files (a file importing node:test) or a project whose tests run under node --test (see node-test:node-test-conventions), TDD workflow/iterations (see vitest:vitest-tdd-workflow), PHP/PHPUnit tests (see phpunit:php-test-conventions).
vitest-tdd-workflow
ACTIVATE when building new features, endpoints, or user stories using TDD with Vitest in a TypeScript/NestJS context. ACTIVATE for '/feature-tdd-dev', 'TDD', 'red-green-refactor' in TS context. Provides Vitest/NestJS-specific TDD examples and commands; cross-language TDD workflow (cross-layer iterations, RED-GREEN-REFACTOR, working app, bug-fix-first-test, mocks-hiding-bugs) lives in craft:tdd-workflow-principles. DO NOT use for: node:test files (a file importing node:test) or a project whose tests run under node --test (see node-test:node-test-conventions), test writing conventions (see vitest:vitest-test-conventions), PHP TDD (see phpunit:php-tdd-workflow).
ts-refactoring
ACTIVATE when refactoring TypeScript code. ACTIVATE for 'refactor', 'extract', 'redesign', 'simplify', or 'clean up' in a TypeScript / Node / NestJS context. This skill provides TS-specific examples for the cross-language refactoring rules defined in craft:refactoring-principles. DO NOT use for: writing new features from scratch, general OOP patterns (see ts-oop).
ts-ports-adapters
ACTIVATE when a TypeScript codebase with no dependency-injection container has to declare a port and wire its adapter — a repository, a clock, an id generator, an event dispatcher the domain depends on. ACTIVATE for 'port', 'adapter', 'SPI', 'dependency inversion' or 'composition root' in a codebase that wires its dependencies by hand. Covers: the port as a type of functions in the domain layer, the adapter as an object literal annotated with it, the handler that takes its dependencies first and its data second, the composition root in the infrastructure layer, and the in-memory stub that substitutes for the adapter in a unit test. DO NOT use for: NestJS, or any container-based DI with decorators and tokens (see nest-ddd-conventions), aggregate operations and smart constructors (see ddd-ts-fp), the Result type itself (see ts-functional).
ts-oop
ACTIVATE when designing TypeScript classes, and the value objects or collections modelled as classes, or when the user asks about object design, encapsulation, Tell Don't Ask, or Symbol.iterator. Provides TS-specific examples for the cross-language OOP rules defined in craft:oop-principles, plus TS-specific patterns (Symbol.iterator, branded types for primitive identifiers). DO NOT use for: a value object or an aggregate modelled as an immutable record rather than a class (see ddd-ts-fp), refactoring methodology (see ts-refactoring).
ts-layer-boundaries
ACTIVATE when wiring a Worker entry point, a Router, a controller, or a domain event consumer in a TypeScript CQRS codebase, or when deciding where a port belongs. ACTIVATE for 'composition root', 'layer boundary', 'ESLint zones', 'domain must not import infrastructure'. Covers: the *Worker.ts composition root, the curried consumer, Router wiring of command endpoints, the five-gesture controller, the domain-to-infrastructure import boundary enforced by ESLint, infrastructure-only ports, in-memory doubles for every port, the read-can-reach-write asymmetry, and read endpoints going through a domain handler. DO NOT use for: the port's own shape and adapter binding (see ts-ports-adapters), the Result type (see ts-functional), domain event structure (see ts-ddd-events).