sergeyitaly
- 34 skills
- 0 followers
- 4 hours ago last updated
- ▌ File Style Conventions · sergeyitalyApply two lightweight file-hygiene conventions when writing or editing files - no emoji characters outside Markdown (.md) files, and YAML files (.yml/.yaml) end with exactly one trailing newline. Use whenever creating or editing non-Markdown files that might contain emoji, or any .yml/.yaml file.
- ▌ Skill Official Updater · sergeyitalyAt the start of a new session, do a cheap check for new or updated official Anthropic skills (github.com/anthropics/skills) and automatically add or update them in skills_library/ (no user prompt). Also use on explicit request ("check for official skill updates", "sync official skills").
- ▌
- ▌ Cross Platform Scripting · sergeyitalyDetect the host OS (Windows/macOS/Linux) and PowerShell version (5.1 Desktop vs 7+ Core) before writing or editing scripts, and write/adapt .ps1, .sh, and .cmd scripts to match what's actually available — avoiding PS7-only syntax on PS5.1, GNU-only flags on macOS/BSD tools, and Windows-only assumptions on POSIX shells. Use before writing a new script, when a script fails with a syntax/parameter error that looks version- or OS-specific, or when asked to make a script cross-platform.
- ▌ Vitest Extension Testing · sergeyitalyRun, debug, and interpret Vitest results in the claude-skills-deployer extension (extension/ directory). Use for unit tests, bench tests, solo tests, and integration tests.
- ▌ Skill Feedback Adaptation · sergeyitalyAUTO-START on new agent session/window (injected by profile-init-watch for Claude, Cursor, Kiro, Copilot) and on new tasks — analyze the prompt and repo, write task-skill-proposals.json, then read top proposed skills before other work. Also register user disagreement into skill-feedback.jsonl when the user says no, not, wrong, stop, or disagrees with agent output.
- ▌ Vscode Extension Publishing · sergeyitalyCreate, package, test, and publish a VS Code extension to the Marketplace using @vscode/vsce. Covers package.json manifest fields, .vscodeignore, local debugging (Extension Development Host), vsce package/publish, version bumps, publisher/PAT setup, and common publish errors. Use when building a new VS Code extension, preparing a release, or debugging `vsce package`/`vsce publish` failures.
- ▌ Cursor Kiro Extension Publishing · sergeyitalyPublish a VS Code-compatible extension to Open VSX for Cursor and Kiro IDE using ovsx. Covers same VSIX as VS Marketplace, OVSX_PAT setup, namespace ownership verification, publish scripts, GitHub Actions, wrong-VSIX pitfalls, and cross-links between registries. Use when publishing to Open VSX, Cursor gallery, Kiro extension registry, or debugging `ovsx publish` failures. Pair with vscode-extension-publishing for packaging and VS Marketplace.
- ▌ CI Preflight · sergeyitalyReproduce a CI pipeline's pre-merge stages (lint, test, validate, build) locally before pushing, by mapping each CI job to its exact local-equivalent command and running them in order. Use when asked to "run CI checks locally", "preflight", "what would fail in the pipeline", or before committing/pushing a change.
- ▌ Profile Init · sergeyitalyAUTO-START when .claude/learning/profile-init-request.json exists with status pending, on a new git branch, or at SessionStart hook injection. Initialize a personal skill profile for the current branch based on team position (DevOps, QA, AQA, Backend, Frontend, BA, Resource Manager, Team Lead). Reads the extension skill catalog, selects skills, writes .claude/profile.local.json (gitignored). Do not wait for the user to ask — run immediately when triggered.
- ▌ Aidlc Tracker · sergeyitalyTrack and advance an AI-DLC (AI-Driven Development Life Cycle) workflow. Reports current phase/stage/status from aidlc-state.md, what's done vs. what's next, reminds about approval gates and audit logging, and updates the tracking docs as stages complete. Use when asked "what's next", "AIDLC status", "what's left", to resume/start a phase, or to record a stage approval.
- ▌ Self Learning · sergeyitalyMaintain a project-local self-learning base of task/command outcomes — record successes and failures with timestamps, durations, and fixes; generate a patterns report (pass rates, recurring errors, known fixes); and surface a learned hint before retrying something that failed before. Use at the start of a session to check learned hints, after running a non-trivial command/skill to record the outcome, when asked "what failed before" or "what did we learn", or to record a manual decision/learning.
- ▌ Skill Creator · sergeyitaly bundleCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
- ▌ Drawio Diagrams · sergeyitalyCreate and edit architecture, network, and infrastructure diagrams as .drawio files using the draw.io MCP server, including Azure architecture diagrams built from the official Microsoft Azure icon set. Use when asked to create, update, or visualize an architecture diagram, infrastructure topology, or system design — especially one involving Azure resources.
- ▌ Adx Schema Check · sergeyitalyCross-check KQL/Kusto queries in application code against the canonical ADX (Azure Data Explorer) table/column schema definitions, and optionally against the live cluster. Flags table/column name mismatches before they cause runtime query failures. Use when working on ADX/Kusto-backed API routes, debugging "table not found"/"column not found" errors, or after changing the schema.
- ▌ Aidlc Doc Writer · sergeyitalyWrite or update AI-DLC methodology documents (inception, per-unit construction, operations, verification) following the standard phase-to-folder mapping and per-document structure conventions, reading the actual source/infra code first so docs reflect what's implemented. Complements aidlc-tracker (status/gates) by producing the document content itself. Use when asked to write/update an AIDLC doc, document a phase/stage, or generate AIDLC scaffolding for a new project.
- ▌ Infra Cost Guard · sergeyitalyEstimate and gate on infrastructure running costs before deploying or leaving resources running — detect expensive resource types in a Terraform plan or live resource list, calculate hourly/daily/monthly estimates, warn before apply, and remind about teardown. Works for Azure, AWS, and GCP. Use before terraform apply on any plan that includes compute, networking (firewalls, NAT gateways, load balancers), or managed databases.
- ▌ CI Pipeline Debug · sergeyitalyDebug a failing CI pipeline stage (lint/test/validate/build/plan/apply/verify or similar) by locating the exact job definition, reproducing its commands locally, and mapping the failure to a root-cause category. Use when a pipeline/job fails and the user wants to know why or wants it fixed.
- ▌ Github Actions CI · sergeyitalyDebug GitHub Actions pipeline failures and reproduce CI stages locally. Use when asked to debug CI, fix a failing workflow, reproduce a job with act, or run a pre-flight check before pushing.
- ▌ Azure Resource Ops · sergeyitalyInvestigate and operate on a project's live Azure resources — build/maintain a resource quick-reference, run common diagnostic commands (health checks, resource state, identity/role checks via az rest, Key Vault secret names, Log Analytics/ADX queries), and record operational quirks so they don't need rediscovering. Use when asked about live resource state, to debug a deployed app, or before running az commands against a project's resources.
- ▌ Gitlab Pipeline Ops · sergeyitalyWork with a GitLab CI/CD project structured like a typical multi-stage pipeline (root .gitlab-ci.yml with stages/workflow/include, per-area job files under .gitlab/ci/*.yml, shared templates and rules in common.yml, manual gates on main for plan/apply/deploy). Use when reading, editing, or debugging .gitlab-ci.yml / .gitlab/ci/*.yml, inspecting pipeline/job status with glab, or working with GitLab CI/CD variables.
- ▌ MCP Server Creation · sergeyitalyBuild, wire, and debug a stdio MCP server bundled inside a VS Code extension (claude-skills-deployer pattern). Covers server code, allow-list security, MCP content-response format, premature-exit fix, deployment to ~/.claude/mcp-servers/, registration in ~/.claude.json for Claude/Cursor/Kiro, auto-start on activation, and health dialog integration. Use when adding a new MCP server or debugging "no output" / early-exit failures.
- ▌ Deployment Practical · sergeyitalyDeployment-first delivery — concrete architecture and IaC over theoretical advice. Use when deploying, provisioning infra, debugging first-apply failures, or when the user wants advice that works on the first attempt (not hand-wavy theory). Pair with Practical Focus toggle (architecture-first / deploy-ready).
- ▌ MCP Efficiency Guide · sergeyitalyRead MCP efficiency hints and fix token-waste patterns. Use when asked about MCP efficiency, token waste, redundant tool calls, or "why is my session so token-heavy". Covers mcp-agent-hints.md, common waste patterns, manual efficiency review, and lazy-mcp.
- ▌ Prompt Context Hooks · sergeyitalyUnderstand how the prompt-context-watch.js hook controls context grounding and practical focus injection. Use when asked "how does context grounding work", "how do I disable prompt injection", "what does practical focus do", "configuring session size warnings", or when modifying this extension's hook system.
- ▌ Skill Usage Insights · sergeyitalyAnalyze recorded skill usage in this project (.claude/learning/runs.jsonl, written by self-learning) and the skills installed in .claude/skills/ to produce a usage and KPI report - which skills are actively used and reliable, which are failing, and which are unused or low-value, with recommendations on what to add or remove. Use when asked for "skill usage stats", "skill KPIs", "which skills should we add or remove", or "are our installed skills still useful".
- ▌ Task Evidence Report · sergeyitalyProduce an evidence-based explanation of what was actually done on a task — real diffs, real test output, real command results, optionally with diagrams — instead of a confident-sounding summary built from memory or assumption. Use when the user needs to explain, defend, present, or hand off completed work to someone else (a reviewer, a manager, a teammate, themselves later), when asked for "proof", "докази", "презентацію що було зроблено", "explain what changed and why", or before closing out a task that involved several commits/approach changes. Explicitly built to survive a skeptical reviewer, not to look finished.
- ▌ Terraform Module Ops · sergeyitalyNavigate a Terraform codebase before changing it — build a module-to-resource map, identify the state backend and provider versions, run the safe local fmt/validate workflow, and flag known-drift resources or operations that need explicit user approval (full apply, destroy, state edits). Use before making Terraform changes, to find which file owns a resource, or to check whether an operation is safe to run.
- ▌ Azure Infra Preflight · sergeyitalyPre-flight checklist before any Azure Terraform deploy — verify az login, detect SSH key type (Azure only accepts RSA), check whether the target resource group already exists and list resources to generate import blocks, validate Terraform version, and write subscription context. Use before terraform plan/apply on a new or potentially pre-existing Azure environment, or when setting up Azure IaC files from scratch.
- ▌ Extension Value Audit · sergeyitalyGive an honest, evidence-based verdict on whether the "Claude Skills Manager" VS Code extension (claude-skill-deployer) is actually delivering real value in THIS project — not a usage-count report of individual skills (see skill-usage-insights for that), but a judgment of the extension's own core mechanisms (adaptive skill suggestions/confidence scoring, task-focus, budget/cost gating, MCP-Force security mode, cross-agent sync, hooks pipeline) against real .claude/learning/ telemetry and the agent's own actual session experience. Use when asked "is this extension helping", "give feedback on Claude Skills Manager", "extension ROI", "does the extension actually do anything", "audit the extension's value", or similar. Produces a blunt, numbers-grounded report — not marketing copy.
- ▌ Terraform Plan Review · sergeyitalyRun terraform fmt/validate/plan and review the output — categorize changes, flag destroys, and triage failures into "real bug" vs "permissions gap" vs "state drift fixable via import block". Use when asked to check Terraform state, review/run a plan, or debug a validate/plan/apply failure.
- ▌ Azure Rbac Diagnostics · sergeyitalyDiagnose Azure RBAC/role-assignment failures (403 AuthorizationFailed, missing identity permissions, Key Vault/ACR/ADX access errors). Distinguishes "Terraform code is correct but the executing identity lacks privilege" from real misconfiguration, and produces the exact az command an admin needs to run. Use when a deployment, terraform apply, or app fails with an authorization/permission error.
- ▌ Cost Attribution Setup · sergeyitalyConfigure and troubleshoot cost attribution — how skill-invoke hooks feed runs.jsonl, what confidence scores mean, how system mode gates optimizers, and when to run Reset Mis-attributed Cost Data. Use when asked "why is my cost low-confidence", "how do I enable cost tracking", "degraded/safe mode", or "reset misattribution".
- ▌ Claude API · sergeyitaly bundleReference for the Claude API and Anthropic SDK (models, pricing, streaming, tool use, MCP, agents, caching, migration). Use when the prompt names Claude/Anthropic or for unstated LLM/agent tasks; see "When to use" below for skip rules.