All authors

Claude Skills by Shehabov
github.com/Shehabov16 skills0 installs1 views
- SkillUse this skill whenever the user wants to create, improve, rewrite, review, tailor, or ATS-optimize a PRODUCT MANAGER resume or CV. Triggers include any mention of 'PM resume', 'product manager resume', 'product resume', 'PM CV', 'product management CV', plain-language asks like 'create my resume', 'build my CV', 'optimize my resume', 'review my resume', 'make it ATS-friendly', 'tailor my resume to this job description', 'help me get hired', the Spanish terms 'hoja de vida' or 'curriculum vit...Votes: 0GitHub stars: 31
- Stack Nextjs SupabaseThe default stack pack, a Next.js App Router app in web/ and a Supabase back end, built and proved through the Supabase MCP from hand-authored migrations, with an offline Postgres proof on PGlite and no Docker or Supabase CLI. Read it by path when PROJECT.md § Stack pack names stack-nextjs-supabase. Use it at kickoff and at run open for the MCP setup and pre-flight, when scaffolding web/ or generating database types and client keys, when writing, reviewing or probing migrations, row level sec...Votes: 0GitHub stars: 4
- Team ArchitectureDesign and record the architecture of the product described in PROJECT.md. Covers the domain model drawn from its Product section, the product invariants from its Product invariants section and the layer that must enforce each one, the architecture of record, API contracts and their conventions, ADRs and where they live, and the task briefs frontend-engineer and backend-engineer build from. Use when making a technical decision, writing or superseding an ADR, issuing a task brief, choosing the...Votes: 0GitHub stars: 4
- Team Brand GuardEnforce the project's brand spec on any surface, string, component or asset. Use before designing, writing copy, building UI or reviewing anything a person will see, when a gate needs a brand check, and whenever a companion design skill's advice meets the brand spec. Covers what to do when the spec is missing or unapproved, the pre-flight checklist, the contrast script, the failures that happen most often with their fixes, the companion skill policy and how to record an override.Votes: 0GitHub stars: 4
- Team Bug RegisterHow bug-historian keeps BUGS.md, the defect register, and uses it to stop the same defect being made twice. Use when a run opens and the team needs its regression brief (stage 1), when the four reviews have passed and the regression guard must check the change against every known defect and every binding standing rule (stage 7), when a run closes and every defect and agent mistake raised in it must be recorded (stage 12), and whenever the Product Lead or a quality role raises a defect. Covers...Votes: 0GitHub stars: 4
- Team Clean CodeThe clean code and commenting standard code-steward enforces at the review-readability gate, so the codebase stays readable for the people and the agents who change it next. Covers naming in the domain's language, the function, file and class thresholds, guard clauses, flag arguments, module headers, docstrings, comments that say why, errors, duplication, tests as specifications and the review checklist with its finding format. Use when writing, reviewing or refactoring any code, and when dec...Votes: 0GitHub stars: 4
- Team Code AnalysisThe line-by-line defect rubric code-analyst applies at the review-defects gate. Covers correctness, line-level security, the data layer, row-level security and database authorisation, concurrency and async, error handling, structural rot measured against fixed thresholds, and the project-specific defects drawn from PROJECT.md, the brand spec and the API contract, with the S1 to S3 finding format, the method and the gate's pass conditions. Use when scanning a diff for bugs, checking complexity...Votes: 0GitHub stars: 4
- Team Code ReviewThe senior review rubric peer-reviewer applies at the review-judgement gate, reading a diff the way a senior engineer reads a colleague's pull request. Covers the reading order, the seven lenses (problem fit, simplicity, boundaries, failure modes, tests, domain language, rollout), the failure-mode catalogue, the comment format, the severity ladder, the verdicts and how each maps to a handoff, and the gate's pass conditions. Use when reviewing a diff or a pull request for judgement before the ...Votes: 0GitHub stars: 4
- Team CopyWrite and review product copy in every locale named in PROJECT.md. Use for any string a person reads, including labels, buttons, errors, empty states, confirmations, status labels, notifications and help text, when reviewing existing copy, and when a locale is added. Covers the writing method and the competitor test, the order locales are written in and the native review rule, the string catalogue and its files, length budgets against the longest locale with pseudo-locale expansion and a leng...Votes: 0GitHub stars: 4
- Team Design SystemThe method for turning the project's design references and brand spec into surfaces that hold at every width, in every locale and in every state. Use when designing, building or auditing any screen, component or flow, including the app shell, navigation, panels, lists, tables, charts, forms and empty states. Covers adopting, adapting or rejecting a reference pattern, shell anatomy, the featured-surface rule, density and rhythm, the five smoothness checks, the card and metric tile rules, the e...Votes: 0GitHub stars: 4
- Team OrchestrationRun planning, right-sizing, gate enforcement, routing and the utilisation check for the team. Use when opening a run (the kickoff interview while PROJECT.md still carries TODO markers, the toolchain pre-flight, writing run.json from the run plan template and expanding its placeholders), when deciding which roles a change needs and recording why each other role is omitted, when dispatching an agent or routing a handoff that came back, when syncing gates and running the utilisation check after ...Votes: 0GitHub stars: 4
- Team ProtocolThe operating loop every agent on the team follows on every task, in five steps (plan, audit the plan, execute, review, hand off). Use at the start of any task an agent performs, before planning or touching a file, and again before writing the handoff. Defines the run folder and its artefacts, the handoff schema and its field rules, the handoff file names for a first pass, a later stage and a fix round, what counts as evidence, how to reject bad input and how to escalate to the Product Lead. ...Votes: 0GitHub stars: 4
- Team ReleaseRelease a change so it can be proved and reversed. Use when release-engineer commits a run's work, runs pre-flight, applies the data layer to the target at release, deploys server functions, tags and pushes, writes the release note, verifies after release or rolls back, and when anyone needs to know whether a change is releasable. Covers who may push, tag or change an environment, the attribution check on every commit, tag and note, branch and commit conventions where PROJECT.md is silent, pr...Votes: 0GitHub stars: 4
- Team SecurityThe security catalogue and sweep security-analyst runs to hold the security gate, on every change, build and commit, and before any release. Seven passes in a fixed order, stopping on a critical in the first two: secrets and keys in the tree and in history; exposure of data stores, storage, endpoints, environment variables and CORS; authentication and access control, including client-side checks and IDOR; injection of every kind and dangerous functions such as eval; dependencies, including pa...Votes: 0GitHub stars: 4
- Team Test ProtocolPlan, run and evidence testing, and decide release readiness. Use when qc-engineer plans or runs a test pass, reproduces or files a defect, or re-tests a fix, when qc-lead audits that evidence, probes by blast radius or writes the readiness report, and whenever someone asks whether a change is tested. Covers the test plan, the test areas (the API contract against the tech-architect's brief, every product invariant in PROJECT.md, guarded state transitions, real flows at every quality-bar width...Votes: 0GitHub stars: 4
- Team Ux AuditAudit a design spec or a built interface against usability heuristics, the accessibility standard in PROJECT.md (default WCAG 2.2 AA), localisation and right to left, the bans in the project's brand spec and the team design system. Use when reviewing a spec before the design gate, auditing a built or shipped surface, checking an implementation against the spec it claims to follow, and when filing, grading or triaging UX findings. Covers the finding format, severities and their gate effect, th...Votes: 0GitHub stars: 4