jq plugins

Scaffold production-grade Claude Code plugins with marketplace integration, validate structure and schemas, audit for security vulnerabilities and best practices, and automate semantic version bumps across manifests and catalogs using auto-invoked skills and interactive commands.

Fully manage Ubiquiti UniFi networks via Rust CLI, TUI, and AI agent for VLANs, SSIDs, firewalls, NAT, DHCP reservations, clients, devices, events, stats, DPI, backups, and raw API access. Audit configurations for security vulnerabilities like open WiFi or permissive rules with ranked remediations. Run health checks on controllers, devices, alarms, and clients. Deploy autonomous agents for diagnostics and ongoing control.

Build, deploy, and operate AI agents on AWS with Amazon Bedrock AgentCore: scaffold projects, connect tools via Gateway and MCP, orchestrate multi-agent systems, enforce Cedar policies, debug with traces and logs, evaluate quality, and harden for production with auth, IAM, rate limiting, and cold-start tuning.

Estimate per-turn token costs for .claude/ configurations and CLAUDE.md files, classifying components as always-loaded, path-scoped, or invoked-only, ranking top contributors, flagging overruns, and fetching exact counts via Anthropic API to optimize Claude context budgets.

Work with 300+ AI models through OpenRouter: query models by price/capabilities, run analytics on usage data, generate images/video/speech, transcribe audio, and integrate LLM calls into TypeScript agents with CLI tools or OAuth authentication—all without managing individual API providers.

Run a multi-lens code review pipeline: deep analysis, automated fix loops, manual promotion of findings, and injection of external issues, all synthesized into a single artifact for PR review.

Delegate tasks to a cloud Devin session with its own VM, shell, browser, and full repo access, enabling parallel work, long-running processes, browser automation, and large refactors.

Coordinate parallel queries to Claude, OpenAI Codex, and Gemini for multi-model code review and problem-solving, then synthesize their responses into a consensus verdict

Query scientific databases like PubMed, arXiv, bioRxiv, medRxiv, ChEMBL, DrugBank, ClinicalTrials.gov, FDA labels, Open Targets, and patents using natural language via Valyu semantic search. Retrieve full-text content, images, compounds, trials, drug data, and relevance scores to accelerate biomedical research, drug discovery, and AI analysis workflows.

Displays real-time context, rate limit status, tool activity, and token costs in the Claude Code status bar, using pure bash and jq.

Generate AI images and videos from text prompts via Leonardo.Ai CLI in your terminal. Specify models, dimensions, enhancements; poll job status, download results to files/directories; list models and check credits.

Daily driver skill pack for Claude Code: read public URLs into clean Markdown, draw hand-drawn diagrams with Excalidraw, build interactive 3D scenes and landing pages, organize GitHub stars by category, count lines across multi-language repos, generate slide decks, configure custom personas, and fetch release notes updates.

Run long-running, unattended batch agents in Claude Code by defining tasks in a CSV table and orchestrating them through a shift-based workflow with self-validation, retries, and subprocess dispatching.

Drive Navaris CLI with AI to fully manage isolated dev sandboxes: create/start/stop/destroy using Incus or Firecracker backends, exec one-shot commands with env/timeout, attach persistent tmux terminals, create/restore snapshots, promote/register images, and handle async operations.

Enforces a team workflow where every change starts with a plan and tests before code, then accumulates those tests into a self-running regression suite with Slack/Discord reporting and cross-repo handoff support.

Generate EN 16931-compliant e-invoices (XRechnung, ZUGFeRD) or plain PDFs by calling the public Scribo HTTP API with curl and jq, no signup or MCP server required.

Plan and implement large features using structured PRDs, SDDs, and ordered task lists. Analyzes codebase to inform planning, then automatically executes tasks with validation against design documents and acceptance criteria.

Operate Crypto.com Exchange via agent-first CLI: fetch real-time market data, execute live trades with dry-runs/safety tiers, manage margins/staking/funding/wallets, analyze portfolios, generate reports, and enable MCP agent autonomy from observation to full trading.

Safely run autonomous Claude Code in production by deploying hooks that block risky Bash commands on unprotected branches, check syntax after edits, monitor context usage, log sessions, and verify setup with diagnostics and guided configuration.

Manage Arize ML observability projects and datasets via ax CLI skills to create from CSV/JSON/Parquet files, list with pagination, retrieve details, delete, export data, and extract IDs using jq for automated platform operations and analytics workflows.

Monitor Claude Code's context usage, 5-hour rate limits, and 7-day usage via color-coded progress bars, plus current folder and git branch, in a cross-platform status line.

Adds a statusline to Claude Code showing git branch/diff, model info, context usage, and rate limits via jq and bash scripts

Define and enforce declarative behavior policies for Claude Code projects, audit configurations against templates, and automate code review, security scanning, and testing through a configurable pipeline of skills, agents, and hooks.

Navigate and analyze codebases semantically with tree-sitter: search symbols by regex, trace call sites and usages, explore class APIs and structures, grep files and contents, scan git commit history, query configs via jq or XPath, and switch workspaces—all locally without LSP setup.

Label a GitHub issue to trigger Zapat's autonomous pipeline powered by Claude Code, which builds, tests, and delivers a merge-ready PR. Setup interactive agent teams, monitor pipeline health with auto-fixes, and add repos for ongoing automation.

Run AI-human collaborative issue tracking in a terminal Kanban TUI (viban). Register problems as prioritized issues, assign to AI agents for automated analysis/implementation/testing in isolated git worktrees, auto-create GitHub PRs, review changes in IDE, approve to merge/cleanup or reject with feedback to iterate.

Enforce a minimum percentage of human-written code by mentoring developers, blocking AI attribution, coaching from session transcripts, and preventing silent data-loss from Timeback/QTI APIs.

Fully automate Wix eCommerce management in Claude Desktop: AI-generate products from images with descriptions/pricing/categories, auto-categorize via analysis, CRUD inventory/orders/customers/events/CMS data, configure taxes/shipping/coupons, segment for targeted emails, build Kanban pipelines tracking carts-to-sales, audit store health, generate revenue analytics and dashboards.

Manage shared, tiered coding memory (canonical/team/user) for architecture decisions, PR notes, and codebase handoffs using Redis/SQL backends. Auto-capture/load memory in Claude sessions via hooks, inspect and sync via CLI/TUI, promote notes, and integrate with git repos for persistent team knowledge.

Store, semantically search, and graph-link persistent project knowledge like decisions, gotchas, patterns, and learnings across Claude sessions. Capture memories from conversations or git changes, check recent work against stored warnings, create multi-agent thinking docs, and monitor memory health with agents and CLI commands.

Convert static web or Next.js projects into ArcBlock blocklets with a provided DID, automate git branch detection and PR creation following repo conventions, set up blocklet dev environments by cloning repos and installing deps, guide onboarding for new developers, and streamline releases via version bumps and changelog updates.

Maintain a persistent knowledge graph for Claude Code projects. Automatically capture lessons learned, architecture decisions, session summaries, and issue tracking from git diffs and chat logs. Retrieve relevant past context during conversations to avoid repeating mistakes and enforce documentation standards.

Trace Konflux container builds from image references back to build logs, source commits, and pipeline URLs using cosign provenance attestations

Validate Claude Code marketplace and plugin directories against OpenPlugins standards with quality scoring, security scanning, schema compliance, and documentation checks, producing prioritized reports and publication readiness assessments.

Manage an Obsidian vault using the PARA methodology: auto-classify inbox items, validate frontmatter, create typed notes (project, area, resource), capture web clippings and voice memos, enrich notes with external data like YouTube transcripts, and git-commit changes automatically.