Emir Kovacevic
Community@Vimurai · Biel
Frontend & AI Product Engineer — building high-performance apps. ~5 years with React, Vue, Svelte, Next & Nuxt, shipping AI-powered product features.
Agent Skills by Emir Kovacevic
Showing 50 vetted skills indexed across 1 GitHub repositories.
Vimuraireview_synthesizer
Aggregates Tier 3 audit stamps from REVIEWS.md and LOG.md into a release readiness verdict.
Vimuraiai-task
Marks completed E-## tasks DONE, runs handover, and surfaces the next task in AI-OS.
Vimuraiai-review-proposed-skills
Review and promote staged self-learned skill proposals through human approval gating.
Vimuraiai-insights
Generates a meta-cognition report from cross-project telemetry stored in a SQLite database.
Vimuraiscope_safety
Enforces filesystem and shell scope boundaries on every file and shell operation.
Vimuraicopilot
Delegates shell command and GitHub CLI lookups to GitHub Copilot CLI.
Vimuraiai-preflight
Bootstraps AI-OS sessions by reading DIGEST, TASKS.md, and stamping SESSION.md.
Vimuraidependency_gate
Documents justification, alternatives, and security review in DECISIONS.md before installing new dependencies.
Vimuraicommit-crafter
Formats and executes Conventional Commits with E-## task IDs and UACS verification stamps.
Vimuraiai-profile
Measures CPU, memory, bundle size, and Web Vitals for Node.js modules inside a sandboxed environment.
Vimuraiai-triage
Analyzes incident logs to detect recurring failures and queue remediation tasks.
Vimuraiux_template
Generates structured UX documentation for UI screens covering layout, states, accessibility, and mobile behavior.
Vimuraiarch-review
Audits codebase structure against the architect blueprint for orphaned work, deviations, and risks.
Vimuraiblueprint-writer
Validates and writes structured architecture blueprints to .ai/blueprints/ before implementation.
Vimuraiai-migration
Guides authoring and validating SQLite schema migration pairs with rollback testing.
Vimuraiai-dispatch
Computes the DAG-aware dispatch frontier of ready versus blocked tasks via orchestrator-mcp.
Vimuraiai-sync-state
Re-reads TASKS.md, architect.md, and DIGEST.md from the filesystem to synchronize cross-agent context.
Vimuraiobs_baseline
Applies structured logging, metrics, tracing, and health check standards to new features and DevOps pipelines.
Vimuraiai-sync-verify
Verify TASKS.md and REVIEWS.md agree with state.sqlite via the task-synchronizer MCP tool.
Vimuraiai-digest
Generates a concise project snapshot by reading AI-OS state files.
Vimuraiai-debug
Enforces a structured hypothesis-test-observe debugging loop with a three-cycle budget for failing tests.
Vimuraici_gate
Documents CI/CD pipeline changes, security implications, and rollback plans before deployment config edits.
Vimurairepo-oracle
Answers historical codebase questions using git log, git blame, and .ai memory files.
Vimuraiai-analyze
Analyzes shell commands for risk and returns PASS, WARN, or BLOCK verdicts before execution.
Frequently Asked Questions About Emir Kovacevic
FAQPage SchemaWhat tasks can I automate using Vimurai's AI-OS skills?▼
You can automate task lifecycle management (marking E-## tasks DONE, handovers), DAG-based dispatch of ready work, tier-aware critic reviews before commits, Conventional Commits formatting, release versioning with signed git tags and CHANGELOG generation, database migrations with rollback paths, and SEO topic cluster creation.
Who are these skills designed for?▼
They target engineers and architects operating a multi-agent AI-OS workflow: an Architect persona handling P-## planning tasks and blueprints, an Engineer persona executing E-## implementation tasks, plus specialized agents like sre_responder, security_engineer, meta_analyst, and performance_engineer for audits, incidents, and profiling.
How do the skills run in practice?▼
Skills are invoked per session via activate_skill (e.g., ai-preflight at session start) or triggered automatically by hooks. They execute through MCP servers — task-synchronizer-mcp, orchestrator-mcp, safe-exec-mcp, code-execution-mcp — with sandboxed allowed-tools like Read, Bash, Grep, and Write, persisting state in .ai/ markdown files and SQLite databases.
What safety and compliance controls are enforced?▼
scope_safety blocks path traversal and unauthorized shell commands on every operation; ai-analyze pre-flights risky commands with PASS/WARN/BLOCK verdicts; dependency_gate and ci_gate require documented justification before changes; identity_guardian audits PII and secrets exposure; commit-crafter forbids author-flag spoofing per the Git Identity mandate.
What prerequisites and dependencies are required?▼
Skills assume an AI-OS project layout with .ai/ files (TASKS.md, DIGEST.md, LOG.md, architect.md), state.sqlite and telemetry.sqlite databases, and running MCP servers (task-synchronizer, orchestrator, safe-exec, archive-manager, advisor). Some capabilities are deferred — e.g., ai-profile uses code-execution-mcp CPU/heap proxies because performance-mcp is not yet built.