Shaharia Lab OÜ
Official@shaharia-lab · Estonia
Innovating at the forefront of AI and technology. We develop solutions to bridge cutting-edge tech with everyday applications, trying to empower everyone
Agent Skills by Shaharia Lab OÜ
Showing 9 vetted skills indexed across 1 GitHub repositories.
shaharia-labui-verify
Screenshot and drive the running Tauri desktop app through WebKit's remote inspector.
shaharia-labapp-demo-ui-screenshot
Generate README screenshots of the Agento desktop app using a synthetic Claude Code session corpus.
shaharia-labarchitect-reviewer
Audits codebases for architectural issues, code smells, and technical debt with severity-ranked findings.
shaharia-labpr-reviewer
Reviews pull requests for correctness, security, performance, architecture, and documentation gaps.
shaharia-lablocal-verify
Reproduce and verify desktop app bugs across backend, browser, and Tauri webview hops.
shaharia-labimplement-issue
Automates GitHub issue implementation from triage through CI-passing, review-clean pull requests.
shaharia-labcontext-updater
Audits and updates documentation and AI context files against recent git changes.
shaharia-labsecurity-reviewer
Audit codebases for vulnerabilities and produce triaged security findings with remediation guidance.
shaharia-labengineering
Implements features, fixes bugs, and refactors code in the Agento Tauri and React codebase.
Frequently Asked Questions About Shaharia Lab OÜ
FAQPage SchemaWhat tasks can I accomplish with Shaharia Lab's skills?▼
You can audit codebases for architecture and technical debt, review pull requests, run security audits with triaged severity findings, verify UI changes by screenshotting and driving the Agento Tauri webview, implement GitHub issues end-to-end into ready PRs, and refresh outdated CLAUDE.md and README documentation.
Who are these skills designed for?▼
Software engineers and maintainers working on the Agento desktop app who need rigorous PR review, security auditing, visual UI verification, and autonomous issue-to-PR delivery. They suit developers who want evidence-based verification rather than assuming a fix works.
How do the review and audit skills run in practice?▼
Skills like pr-reviewer, architect-reviewer, and security-reviewer run in a forked context as general-purpose agents on the opus model, using Read, Grep, Glob, Bash, and Task. You pass an argument such as a PR number, branch, or audit scope like 'last 7 days changes'.
What prerequisites or environment do these skills require?▼
UI verification and demo screenshot skills require the running Agento desktop app with its real Tauri webview; demo screenshots use a synthetic anonymised dataset without touching real ~/.agento or ~/.claude data. Issue implementation requires GitHub access, git worktrees, and CI.
Can these skills keep project documentation accurate?▼
Yes. The context-updater skill identifies and surgically updates outdated CLAUDE.md, README.md, and docs/ files after feature work or PRs. You can scope it by time range, such as 'since last 7 days', or run a full documentation check.