agent_tools

macOS clipboard intelligence CLI for agents and power users. Use when working with the system clipboard, generating formatted Excel tables from CSV or JSON, editing clipboard table cells, building or editing nested HTML/Markdown lists, managing reusable HTML templates, or converting between RTF, HTML, plain text, and Jinja2. Triggers on clipboard read/write/inspect operations, Excel-pasteable HTML/SVG/PNG table generation, nested list generation/editing, reusable template capture/render/paste, privacy-aware clipboard history/watch/filter/prune/restore workflows, previewing rendered HTML without touching the clipboard, shell completions, doctor readiness checks, and agent-command templatization or batch rendering. Requires macOS. Binary must be built first with cargo build --release.

Use xli for deterministic, JSON-first Excel workbook operations: inspect, read, write, format, validate, template, batch edits, sheet management, and workbook creation from CSV/Markdown/JSON. Trigger on `.xlsx` automation, report generation, workbook validation, table import, formula edits, or atomic spreadsheet mutations.

Manage a file-based knowledge vault with YAML frontmatter, sidecar markdown, and JSONL indexing. Use when initializing or maintaining a knowledge base, attaching metadata to markdown or non-markdown assets, rebuilding or validating INDEX.jsonl, or searching vault records by metadata. Trigger on: "knowledge base", "vault", "frontmatter", "sidecar", "INDEX.jsonl", "document this query/template/skill", "search the vault", or "set up KB docs".

Use Jirali to inspect, create, transition, reconcile, report on, and audit Jira work from an agent-safe CLI.

Use notionli for agent-safe Notion workspace operations: search, page reads, block edits, database/data-source inspection, row upserts, comments, users, files, snapshots, sync, and auditable dry-run writes. Trigger on Notion automation, schema discovery, database row edits, page/block mutations, or any task that needs structured Notion access.

Operate the lira CLI, a local agent-native task system where local tickets are local-first issue records decomposed into acceptance criteria plus atomic tasks. Use when the user wants to break down Jira/GitHub/local work into agent tickets; create, claim, work, comment on, sync, query, or close local tickets; manage acceptance criteria and atomic tasks; resolve GitHub sync conflicts; or expose local candidates and normalized issue projections for a Symphony-style runner through ~/.lira/.

Use barli to build and operate a macOS menu bar automation app that discovers Python functions via decorators and configures menus from YAML. Use when the user wants local menubar actions, hot-reloaded Python plugins, YAML-driven command menus, or lightweight GUI launchers for agent tools.

Use bashli to execute structured shell workflows for AI agents using JSON or YAML task specs instead of raw bash strings. Use when the user needs safe command orchestration, variable capture, parallel read-only steps, assertions, output transforms, token-aware truncation, or structured command results.

Use docli to inspect, create, patch, query, render, validate, and finalize Word .docx documents from an agent-native CLI. Use when the user needs structured DOCX automation, high-fidelity document edits, OOXML-safe patching, templates, comments, redlines, or JSON-readable document operations.

Use framerli to inspect, publish, promote, and manage Framer projects through an agent-safe Rust CLI with a Node bridge to the official framer-api SDK. Use when the user asks for Framer project status, CMS collections/items, deployment promotion, publishing, or Framer API automation.

Use vizli to discover visualization templates, resolve parameters, render charts/diagrams/explainers, and verify generated visual artifacts. Use when the user asks for template-driven SVG/PNG/HTML visualizations, explainers, diagrams, or repeatable data-to-visual workflows.

Use mdli to read, edit, and regenerate structured Markdown documents as an AST. It owns sections, tables, frontmatter, managed blocks, and recipe-driven apply for agent workflows that would otherwise hand-build .md files.

Create interactive visual explainers — HTML files with embedded SVG diagrams, charts, sliders, and steppers that teach concepts through spatial and interactive means. Use this skill whenever the user asks to "explain", "visualize", "diagram", "illustrate", or "show how something works", and the output should be a standalone file they can open in a browser. Also trigger when the user asks for an interactive demo, a visual walkthrough, an animated explanation, a calculator or simulator for a concept, or any educational/explanatory artifact that benefits from interactivity rather than static text. This includes statistical concepts, system architectures, algorithms, scientific processes, business models, decision frameworks, data pipelines, and any "how does X work" question where seeing is understanding. Do NOT use for slide decks (use pptx), formal reports (use docx), or data dashboards with real data connections. This skill is for teaching and explaining through interactive visuals.

Operate the vizli tool to turn existing vizli templates into charts, diagrams, explainers, and other rendered artifacts. Use this skill when the user asks to render, generate, preview, inspect, validate, or troubleshoot a visualization with vizli, or when they want a chart or diagram produced from an existing vizli template library and data files. This skill is for template discovery, payload resolution, command execution, and output verification. It is not for authoring new vizli templates from scratch unless the user explicitly asks for conversion or template creation.

Create, edit, and improve AGENTS.md files for Codex projects. Use this skill when asked to write, audit, refactor, or optimize a AGENTS.md (or Codex.local.md) file. Produces files that make Codex sessions dramatically more effective by encoding institutional knowledge, project conventions, and critical context directly into the project's memory layer.

Use this skill whenever you need to create, read, update, close, comment on, label, assign, milestone, pin, lock, transfer, or otherwise manage GitHub Issues using the official GitHub CLI (gh). Covers the full issue lifecycle including search, bulk operations, branch development from issues, GitHub Projects integration, and the REST API escape hatch for anything gh issue subcommands don't expose directly.

Load this skill when making any scrum or agile judgment call: sprint planning, backlog refinement, story sizing, sprint review, retrospectives, blocker triage, definition of done assessment, or any situation where the right scrum move is unclear. This is a reference skill — no tools, only doctrine and decision frameworks calibrated for an agentic context.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.

Create, edit, and improve CLAUDE.md files for Claude Code projects. Use this skill when asked to write, audit, refactor, or optimize a CLAUDE.md (or CLAUDE.local.md) file. Produces files that make Claude Code sessions dramatically more effective by encoding institutional knowledge, project conventions, and critical context directly into the project's memory layer.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.