skill

agent-skills

A curated collection of custom AgentSkills and token-saving MCP servers for extending the capabilities of AI coding assistants like Claude Code.

kambleakash0/agent-skillscommunitylistedunreviewednot running here yet

Use skill

Set up

Skills are installed by placing their folder where your agent loads skills.

Install instructions come from the README

Skills do not use an MCP config. Follow the steps in the README below; one-click install arrives with accounts.

README

synced Oct 14, 2026

agent-skills

A personal collection of Claude Code agent skills and MCP servers, can be used with other agents too.

Available Skills

Skill	Command	Description
git-workflow	`/git-workflow`	Guided Git workflow: branching, commits, PRs, and conflict resolution
code-review	`/code-review`	Thorough code review covering correctness, security, performance, and style
english-humanizer	`/humanize`	Humanize the text
grill-master	`/grill`, `/grill-master`, `/grill-me`	Get relentlessly interviewed about a topic, plan or design until every branch of the decision tree is resolved.
spec-writer	`/spec`, `/spec-writer`	Create a PRD through user interview, codebase exploration, and module design, then submit as a GitHub issue. Use when user wants to write a PRD, create a product requirements document, or plan a new feature.
slice-the-spec	`/slice`, `/slice-the-spec`	Break a PRD into independently-grabbable GitHub issues using vertical slices.
incremental-tdd	`/tdd`, `/incremental-tdd`	Test-driven development with a red-green-refactor loop. Builds features or fixes bugs one vertical slice at a time.
deep-codebase-audit	`/deep-codebase-audit`, `/deep-audit`, `/codebase-audit`	Identify architectural friction in a codebase and refactor shallow, tightly coupled modules into deeper, more cohesive ones—improving testability, structure, and making it easier (and safer) for AI agents to understand and modify.
spec-to-plan	`/spec-plan`, `/spec-to-plan`	Turn a PRD into a multi-phase implementation plan using tracer-bullet vertical slices.
domain-glossary	`/domain-glossary`, `/glossary`	Turn an ongoing conversation into a DDD-style ubiquitous language document. It scans for domain terms, resolves ambiguities, proposes canonical names, and writes a living glossary to `DOMAIN_GLOSSARY.md` in the working directory.
script-writer	`/script-writer`, `/write-a-script`	Script-writer that drafts presentations, essays, emails, and slides using only the cognitive and persuasive heuristics from Patrick Winston's "How to Speak" lecture.
teach-me	`/teach-me`	Teach the user a new skill or concept within this workspace. Stateful — the user's learning state persists across sessions and is tracked in workspace files.
context-pack	`/context-pack`, `/handoff`	Compact the current conversation into a handoff document so a fresh agent can continue the work in the next session.

MCP Servers

Server	Description
ast-editor	AST-targeted code editing MCP server with 28 surgical tools: functions, classes, methods, fields, parameters, imports, parametrized tools for leading comments (`edit_leading_comment(op=...)`), sibling inserts (`insert_sibling`), body inserts (`insert_in_body(at/after/before)`), and tiered reads (`read_symbol(depth=full/interface/signature)`). In-body snippet editing (`replace_in_body`, `delete_in_body`), dict/list editing across JSON/YAML/TOML and Python/JS/TS literals. Dotted targets descend into closures (Go `stdioCmd.RunE`, TS `app.handler`). Built on tree-sitter. Supports 11 languages: Python, JavaScript, TypeScript, C, C++, Ruby, Go, Java, JSON, YAML, and TOML.
notebook-editor	Jupyter notebook editing MCP server with 23 cell-level tools: notebook creation (`create_notebook` with valid nbformat schema), cell structure (add/delete/move/split/merge), content editing (replace/prepend/append), outputs & metadata management, notebook-wide symbol discovery, and kernel-agnostic execution (execute_cell, execute_all_cells, restart/interrupt/shutdown kernel — works with any installed Jupyter kernel: python3, ir, julia-1.x, ...). Complements `ast-editor` — use this for `.ipynb` files, `ast-editor` for everything else.
rmd-editor	R Markdown editing MCP server with 24 chunk- and prose-level tools: document creation (`create_rmd`), cell structure (add/delete/move/split/merge), content editing (replace/prepend/append), chunk options (get/set/remove with a special `label` key), YAML frontmatter (get/set/remove), text search, and R kernel execution via IRkernel (outputs returned inline, never persisted — R Markdown regenerates on render). Complements `notebook-editor` — use this for `.rmd` files, `notebook-editor` for `.ipynb`.

Skills installation via skills.sh

Install multiple skills at once:

npx skills@latest add kambleakash0/agent-skills

or, install individual skills:

npx skills@latest add kambleakash0/agent-skills/skills/git-workflow
npx skills@latest add kambleakash0/agent-skills/skills/code-review
npx skills@latest add kambleakash0/agent-skills/skills/english-humanizer
npx skills@latest add kambleakash0/agent-skills/skills/grill-master
npx skills@latest add kambleakash0/agent-skills/skills/spec-writer
npx skills@latest add kambleakash0/agent-skills/skills/slice-the-spec
npx skills@latest add kambleakash0/agent-skills/skills/incremental-tdd
npx skills@latest add kambleakash0/agent-skills/skills/deep-codebase-audit
npx skills@latest add kambleakash0/agent-skills/skills/spec-to-plan
npx skills@latest add kambleakash0/agent-skills/skills/domain-glossary
npx skills@latest add kambleakash0/agent-skills/skills/script-writer
npx skills@latest add kambleakash0/agent-skills/skills/teach-me
npx skills@latest add kambleakash0/agent-skills/skills/context-pack

Skill Format

Each skill is a Markdown file with a YAML front-matter block followed by the prompt body, following the skills.sh spec:

---
name: skill-name
description: >
  What the skill does and when to trigger it.
  TRIGGER when the user writes /skill-name.
metadata:
  author: kambleakash0
  version: x.y.z
triggers:
  - /trigger-1
  - /trigger-2
  - ...
---

Skill prompt body goes here…

Adding New Skills

Create a new folder in the root of skills/. The folder name must strictly match your skill's name in the YAML frontmatter.
Add a SKILL.md file in the root of the new folder. Ensure it contains a name (max 64 chars) and description (non-empty, max 1024 chars) inside the YAML frontmatter.
If needed, add a resources/ folder in the root of the new folder and populate it with the necessary files.
Add a row to the Available Skills table in this README.
Test it locally by placing the folder from skills/ in ~/.claude/skills/ or ~/.agents/skills/.

MCP Servers' installation

Refer to their respective READMEs for installation instructions.