Aman Pal
Community@amanpal3 · Lucknow ,India
Building at the intersection of AI& Web Development |Python | Flask | JavaScript | REST APIs| Full-stack Dev.AI/ML Student · Open source Enthusiast ·Hackathon
Agent Skills by Aman Pal
Showing 57 vetted skills indexed across 1 GitHub repositories.
amanpal3code-review
Reviews code for bugs, security vulnerabilities, readability, and performance issues.
amanpal3ponytail
Enforces minimal, YAGNI-driven code changes using stdlib-first solutions and shortest working diffs.
amanpal3ponytail-audit
Audits an entire repository for over-engineering and ranks code to delete or simplify.
amanpal3ponytail-help
Displays a reference card of ponytail modes, skills, and commands.
amanpal3ponytail-review
Reviews diffs for over-engineering and lists code to delete or simplify.
amanpal3ponytail-gain
Displays a benchmark scoreboard comparing code size, cost, and speed with and without ponytail.
amanpal3ponytail-debt
Scans repositories for ponytail shortcut comments and compiles them into a technical debt ledger.
amanpal3commit
Write conventional git commit messages with type, scope, and subject.
amanpal3prd
Generate structured JSON Product Requirements Documents with stories and quality gates.
amanpal3dev-browser
Automates browser interactions with persistent page state using Playwright and CDP.
amanpal3ralph-loop
Executes autonomous build loops that implement user stories from a PRD JSON with quality gates and atomic commits.
amanpal3codebase-mapper
Analyzes codebase structure, dependencies, patterns, and technical debt to generate ARCHITECTURE.md and STACK.md.
amanpal3context-health-monitor
Detects context degradation signals and writes state dumps to .gsd/STATE.md before sessions fail.
amanpal3subagent-delegation
Delegates GSD workflow tasks to Antigravity subagents with workspace isolation and inline fallback.
amanpal3planner
Creates executable phase plans with task breakdown, dependency graphs, and goal-backward verification.
amanpal3context-fetch
Searches codebases with keywords before reading targeted file sections to reduce context usage.
amanpal3context-compressor
Compresses file context into summaries, outlines, and diffs to reduce token usage.
amanpal3debugger
Diagnose software bugs through systematic hypothesis testing and persistent state tracking.
amanpal3empirical-validation
Validates completed work with empirical evidence before marking tasks done.
amanpal3verifier
Validates implemented code against phase requirements with empirical evidence checks.
amanpal3plan-checker
Validates PLAN.md files against six quality dimensions before execution.
amanpal3executor
Executes GSD plans with atomic commits, deviation handling, and checkpoint protocols.
amanpal3token-budget
Estimates and tracks token usage to prevent LLM context window overflow.
amanpal3gsd-context-health-monitor
Detects context degradation signals and writes state dumps to .gsd/STATE.md.
Frequently Asked Questions About Aman Pal
FAQPage SchemaWhat tasks can I accomplish with amanpal3's skill registry?▼
You can run spec-driven development via the GSD methodology, review and refactor code, write conventional commits, generate PRDs, architect backends with REST/GraphQL/gRPC, build unit-to-E2E test suites, apply OWASP security practices, automate browsers, design UI/UX, and optimize pages for SEO.
Who is the target audience for these skills?▼
Full-stack developers, backend engineers, and AI-assisted coding practitioners. The bio indicates Python, Flask, JavaScript, and REST API focus, while skills like codebase-mapper, debugger, executor, and testing-code serve engineers managing planning, implementation, verification, and technical debt.
How do the GSD and ponytail skills work together in practice?▼
GSD enforces spec-driven planning, state persistence, and empirical verification through planner, executor, and verifier skills. Ponytail complements it by auditing diffs and repos for over-engineering, tracking shortcut debt, and enforcing YAGNI and stdlib-first decisions during reviews.
Are these skills open source and what do they cost?▼
The ponytail skill family (ponytail, ponytail-audit, ponytail-review, ponytail-help, ponytail-gain, ponytail-debt) is explicitly MIT licensed via its GitHub homepage. Other skills in the manifest do not declare licenses, so their terms should be confirmed per repository before reuse.
What prerequisites or integrations do these skills require?▼
antigravity-agents requires bash, file editing, and web search capabilities plus Claude, Gemini, or OpenAI models. Other skills integrate with Jira, Slack, Notion, Firebase, Sentry, Google Stitch, Codex CLI, and testing frameworks like Vitest, Jest, Pytest, and Playwright.