Skip to content

Commit 2ef97af

Browse files
antonyevansclaude
antonyevans
and
claude
committed
feat: improve repo discoverability — root README, skills-index.json, domain READMEs
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
1 parent eef57e9 commit 2ef97af

8 files changed

Lines changed: 330 additions & 51 deletions

File tree

‎README.md‎

Lines changed: 178 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,179 @@
1-
# skill-engineer
2-
Shared public skills developed by Skill Engineer
1+
# AI Agent Skills Library – Executable Workflows for LLMs
32

4-
Get your own at: https://agentengineermaster.com/skill-engineer
3+
A curated, production-grade library of structured skills for AI agents. Drop any skill into your Claude Code project and invoke it immediately. Built and maintained by [Skill Engineer](https://agentengineermaster.com).
4+
5+
---
6+
7+
## What This Is
8+
9+
This repository contains structured, executable skills for AI agents — self-contained workflow definitions that any Claude Code instance can load and run. Each skill encodes a repeatable, expert-level process as a machine-readable SKILL.md file: triggers, inputs, step-by-step execution logic, outputs, and validation rules.
10+
11+
If you work with AI agents, LLM workflows, or prompt engineering, these skills save you from writing complex context from scratch. Instead of prompting from zero, you load a skill and invoke a proven workflow. Think of this library as reusable agent workflows — the equivalent of a shared component library, but for AI reasoning and execution patterns.
12+
13+
Keywords this library covers: AI agents, LLM workflows, prompt engineering, Claude skills, Copilot skills, reusable agent workflows, context engineering, autonomous agents, workflow automation.
14+
15+
---
16+
17+
## How Skills Work
18+
19+
Each skill lives in a subfolder inside its domain directory. The entry point is always `SKILL.md`.
20+
21+
### SKILL.md Structure
22+
23+
**Frontmatter (YAML):**
24+
25+
```yaml
26+
---
27+
name: skill-folder-name
28+
description: One-sentence description of what the skill does and when to trigger it.
29+
argument-hint: "[optional argument format]"
30+
triggers:
31+
- "trigger phrase 1"
32+
- "/slash-command"
33+
---
34+
```
35+
36+
**Body sections:**
37+
38+
| Section | Purpose |
39+
|---|---|
40+
| When to use | Exact conditions that should trigger this skill (and explicit do-not-use cases) |
41+
| Inputs | Required and optional parameters |
42+
| Steps | Numbered execution phases — what the agent does, in order |
43+
| Outputs | What gets produced: files, reports, structured data |
44+
| Validation | How the agent self-checks its output before returning |
45+
| Rules | Hard constraints the agent must not violate |
46+
47+
Each skill is self-contained and loadable by any Claude Code instance. No dependencies on external services unless explicitly documented in the skill.
48+
49+
---
50+
51+
## Skill Domains
52+
53+
### design/
54+
Presentation and visual asset creation skills for designers and creative teams.
55+
56+
| Skill | Description | When to Use |
57+
|---|---|---|
58+
| [marp-slides](design/marp-slides/) | Creates MARP presentation decks (.md files rendered to PDF/HTML/PPTX via marp CLI) with custom CSS themes, SVG inline charts, dashboard components, and speaker notes | Creating slides, building a deck, making a presentation, generating MARP output, editing existing slides |
59+
60+
---
61+
62+
### engineering/
63+
Code quality, security, and debugging skills for software engineers and technical teams.
64+
65+
| Skill | Description | When to Use |
66+
|---|---|---|
67+
| [fix-from-logs](engineering/fix-from-logs/) | Diagnoses bugs from raw error logs, stack traces, or CI failure output — triages and clusters errors, localizes root cause, proposes a fix + regression test | User pastes logs and wants the underlying bug fixed with tests |
68+
| [security-mindset-master](engineering/security-mindset-master/) | Gates implementation of API endpoints, auth logic, database queries, and user input handling with threat surface analysis, secure defaults verification, and attacker's eye pass | Implementing or modifying any feature that stores or transmits user data |
69+
70+
---
71+
72+
### legal/
73+
Compliance and regulatory assessment skills for legal and risk teams.
74+
75+
| Skill | Description | When to Use |
76+
|---|---|---|
77+
| [compliance-check](legal/compliance-check/) | Assesses US regulatory compliance likelihood for any consumer product before listing for US sale — classifies product, determines applicable standards, produces a structured verdict | Before listing any consumer product for US sale |
78+
79+
---
80+
81+
### marketing/
82+
Go-to-market, brand, and competitive analysis skills for marketing teams.
83+
84+
| Skill | Description | When to Use |
85+
|---|---|---|
86+
| [analyzing-dtc-stores](marketing/analyzing-dtc-stores/) | Produces an investor-grade teardown of a DTC brand from its public URL, covering brand, market, unit economics, supply chain, channel mix, marketing, and agentic-commerce readiness | DTC teardown, brand teardown, competitor analysis, investor memo |
87+
| [audit-brand](marketing/audit-brand/) | Audits and evolves brand positioning, voice consistency, and messaging for an existing brand — runs a structured audit and produces a scorecard with fix recommendations | Brand health checks, positioning refreshes, messaging framework development |
88+
| [marketing-plan](marketing/marketing-plan/) | Develops a full go-to-market marketing plan covering ICP, competitive landscape, channel strategy, budget allocation, pricing, content strategy, and 90-day launch timeline | "Build a marketing plan", "create a GTM strategy", planning a product launch |
89+
90+
---
91+
92+
### operations/
93+
Strategic analysis, planning, and research optimization skills for operations teams.
94+
95+
| Skill | Description | When to Use |
96+
|---|---|---|
97+
| [auto-research](operations/auto-research/) | Autonomously optimizes any Claude skill (SKILL.md) or CLAUDE.md file through a closed hypothesis→test→evaluate→keep/discard loop | Improving a skill's output quality against measurable criteria |
98+
| [morning-brief](operations/morning-brief/) | Generates a structured morning brief from current priorities, active task board, competitor digest, and content pipeline status | Daily morning planning, intention setting |
99+
| [stress-test](operations/stress-test/) | Three-phase strategic decision analysis combining verbalized sampling, customizable analytical lenses, and a structured decision brief — surfaces non-obvious tail-distribution insights | Any high-stakes decision: pivots, category selection, positioning, pricing, hiring |
100+
101+
---
102+
103+
### product/
104+
User research, feedback analysis, and idea validation skills for product teams.
105+
106+
| Skill | Description | When to Use |
107+
|---|---|---|
108+
| [analyzing-feedback](product/analyzing-feedback/) | Parses raw customer feedback into categorized severity-ranked themes, scans the codebase to identify affected files, and proposes 3 structural edit options per theme | Translating customer complaints into file-level engineering proposals |
109+
| [find-my-business](product/find-my-business/) | Guides a founder from "no idea" through idea generation, rapid validation, and commitment to a business — synthesizing PG, YC, Lean Startup, Mom Test, and JTBD methodology | Finding a business to start, exploring startup ideas, getting unstuck on what to build |
110+
111+
---
112+
113+
## Example Skill (Inline): `operations/stress-test`
114+
115+
**Trigger:** `/stress-test [decision or choice to analyse]`
116+
117+
**What it does:** Runs a three-phase strategic decision analysis. Phase 1 uses verbalized sampling across four analytical lenses (quantitative, strategic, risk-first, contrarian) to surface non-obvious insights that mode-collapsed prompting suppresses. Phase 2 synthesizes into a decision brief with a recommendation and confidence level. Phase 3 runs a critic pass to stress-test the brief.
118+
119+
**Example usage:**
120+
```
121+
/stress-test should I enter the home wellness category or stay narrower?
122+
/stress-test hire a contractor now vs wait until revenue hits $10k/mo
123+
/stress-test two content formats: long-form carousel vs short-form daily posts
124+
```
125+
126+
**Output format:** Structured markdown decision brief — verbalized sampling outputs per perspective, synthesis narrative, ranked options with rationale, recommendation with confidence level, and critic objections addressed.
127+
128+
---
129+
130+
## How to Use
131+
132+
### 1. Developer: copy a skill into your project
133+
134+
Copy any skill folder into `.claude/skills/` in your project:
135+
136+
```bash
137+
cp -r operations/stress-test /your-project/.claude/skills/stress-test
138+
```
139+
140+
Then invoke it in Claude Code:
141+
142+
```
143+
/stress-test should I build this feature in-house or use a third-party API?
144+
```
145+
146+
Claude Code auto-discovers SKILL.md files in `.claude/skills/` and registers them as slash commands. No configuration required.
147+
148+
### 2. AI agent platform: load as a tool definition
149+
150+
Each SKILL.md is structured to load cleanly as a tool definition for any LLM-based agent platform:
151+
152+
- Parse the YAML frontmatter for `name`, `description`, `triggers`, and `argument-hint`
153+
- Load the body as the tool's system prompt or instruction block
154+
- Use `skills-index.json` at the repo root for programmatic discovery — it contains every skill's name, path, domain, description, and tags in a machine-readable array
155+
156+
```json
157+
// skills-index.json entry example
158+
{
159+
"name": "stress-test",
160+
"path": "operations/stress-test",
161+
"domain": "operations",
162+
"description": "Three-phase strategic decision analysis...",
163+
"tags": ["decision-analysis", "strategy", "verbalized-sampling", "founder", "risk-assessment"]
164+
}
165+
```
166+
167+
---
168+
169+
## Get a Custom Skill Built
170+
171+
Need a skill that isn't in this library? [Skill Engineer](https://agentengineermaster.com) builds custom skills to the same production standard — structured, tested, and ready to drop into any Claude Code project.
172+
173+
Commission a custom skill at **https://agentengineermaster.com**
174+
175+
---
176+
177+
## Keywords
178+
179+
This library covers AI agent skills, LLM workflows, prompt engineering framework, reusable agent workflows, and context engineering patterns for production use. It includes Claude Code skills structured for immediate slash-command invocation, GitHub Copilot skills loadable via tool definitions, and autonomous AI agents workflows spanning marketing, engineering, product, legal, design, and operations. Use this library to accelerate workflow automation with proven, expert-encoded skill definitions — a structured skill library built for any team deploying AI agents at scale.

‎design/README.md‎

Lines changed: 10 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -1,15 +1,17 @@
11
# Design
22

3-
Skills for designers and creative teams.
3+
Skills for designers and creative teams who need to produce polished presentations and visual assets using AI agents.
44

5-
## Skills in this folder
5+
## Skills
66

7-
- UI/UX design and critique
8-
- Brand identity and guidelines
9-
- Design system documentation
10-
- User flow and wireframe analysis
11-
- Visual QA and accessibility review
7+
| Skill | Description | When to Use |
8+
|---|---|---|
9+
| [marp-slides](marp-slides/) | Creates MARP presentation decks (.md files rendered to PDF/HTML/PPTX via marp CLI) with custom CSS themes, SVG inline charts, dashboard components, and speaker notes | Creating slides, building a deck, making a presentation, generating MARP output, editing existing slides, converting content into slide format |
10+
11+
## When to use this domain
12+
13+
Use design skills when you need to produce visual outputs from structured content — presentations, reports, or branded documents. These skills handle the rendering pipeline and CSS theming so you can focus on content, not formatting. They are not for standalone image generation or design critique workflows.
1214

1315
---
1416

15-
*Part of the [Skill Engineer Master](https://www.agentengineermaster.com/skill-engineer) shared skills library.*
17+
*Part of the [Skill Engineer](https://agentengineermaster.com) shared skills library.*

‎engineering/README.md‎

Lines changed: 11 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -1,15 +1,18 @@
11
# Engineering
22

3-
Skills for software engineers and technical teams.
3+
Skills for software engineers and technical teams who need AI agents to diagnose bugs, fix code, and enforce security standards during implementation.
44

5-
## Skills in this folder
5+
## Skills
66

7-
- Debugging and root cause analysis
8-
- Security implementation and threat modeling
9-
- Code quality, testing, and TDD
10-
- Architecture and system design
11-
- CI/CD, deployment, and observability
7+
| Skill | Description | When to Use |
8+
|---|---|---|
9+
| [fix-from-logs](fix-from-logs/) | Diagnoses bugs from raw error logs, stack traces, or CI failure output — triages and clusters errors, localizes root cause, proposes a targeted fix and writes a typed pytest regression test | User pastes logs wanting the underlying bug fixed with tests that prevent recurrence; CI output shows failures needing code changes |
10+
| [security-mindset-master](security-mindset-master/) | Gates implementation of API endpoints, auth logic, database queries, and user input handling with threat surface analysis, secure defaults verification, and attacker's eye pass before any code ships | Implementing or modifying any API endpoint, authentication or authorization logic, database queries, session/token management, file uploads, or any feature that stores or transmits user data |
11+
12+
## When to use this domain
13+
14+
Use engineering skills when you need an AI agent to act as a specialist during implementation — not as a general coding assistant. These skills enforce a structured process before code is written or merged, catching bugs and security issues at the right phase. They are not for code review on already-shipped code or architectural design without a concrete change in scope.
1215

1316
---
1417

15-
*Part of the [Skill Engineer Master](https://www.agentengineermaster.com/skill-engineer) shared skills library.*
18+
*Part of the [Skill Engineer](https://agentengineermaster.com) shared skills library.*

‎legal/README.md‎

Lines changed: 10 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -1,15 +1,17 @@
11
# Legal
22

3-
Skills for legal, compliance, and risk teams.
3+
Skills for legal, compliance, and risk teams who need AI agents to assess regulatory exposure before taking action.
44

5-
## Skills in this folder
5+
## Skills
66

7-
- Contract review and red-lining
8-
- Regulatory compliance assessment
9-
- Privacy and data handling policies
10-
- IP and trademark research
11-
- Risk identification and mitigation
7+
| Skill | Description | When to Use |
8+
|---|---|---|
9+
| [compliance-check](compliance-check/) | Assesses US regulatory compliance likelihood for any consumer product before listing for US sale — classifies product into a regulatory category, determines applicable standards, gathers market evidence, and produces a structured verdict (green/yellow/orange/red) | Before listing any consumer product for US sale; sourcing decisions where regulatory risk is unknown; pre-launch compliance screening |
10+
11+
## When to use this domain
12+
13+
Use legal skills when you need a structured compliance or regulatory check before committing to a decision — particularly for product sourcing, listing, or launch where the wrong call carries legal or financial risk. These skills are not a substitute for qualified legal advice and do not produce legal opinions. They are designed for pre-screening and evidence gathering at speed.
1214

1315
---
1416

15-
*Part of the [Skill Engineer Master](https://www.agentengineermaster.com/skill-engineer) shared skills library.*
17+
*Part of the [Skill Engineer](https://agentengineermaster.com) shared skills library.*

‎marketing/README.md‎

Lines changed: 12 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -1,15 +1,19 @@
11
# Marketing
22

3-
Skills for marketing teams and content creators.
3+
Skills for marketing teams, founders, and content creators who need AI agents to run structured go-to-market, brand, and competitive analysis workflows.
44

5-
## Skills in this folder
5+
## Skills
66

7-
- Content creation (blog, social, email)
8-
- SEO research and optimisation
9-
- Campaign planning and copy
10-
- Brand voice and messaging
11-
- Analytics and performance reporting
7+
| Skill | Description | When to Use |
8+
|---|---|---|
9+
| [analyzing-dtc-stores](analyzing-dtc-stores/) | Produces an investor-grade teardown of a DTC brand from its public URL — covering brand, market, unit economics, supply chain, channel mix, marketing, and agentic-commerce readiness | DTC teardown, competitor analysis, brand breakdown, investor memo on any ecommerce brand |
10+
| [audit-brand](audit-brand/) | Audits and evolves brand positioning, voice consistency, and messaging for an existing brand — produces a scorecard with specific fix recommendations and reusable messaging frameworks | Brand health checks, positioning refreshes, messaging framework development for personal brands and DTC stores |
11+
| [marketing-plan](marketing-plan/) | Develops a full go-to-market marketing plan covering ICP, competitive landscape, channel strategy with buyer journey mapping, budget allocation, pricing, content strategy, 90-day launch timeline, and AARRR + GEV KPI framework | "Build a marketing plan", "create a GTM strategy", planning the launch for a new product or brand |
12+
13+
## When to use this domain
14+
15+
Use marketing skills when you need a structured, repeatable process for go-to-market work — not one-off content creation or ad-hoc research. These skills produce durable outputs (plans, scorecards, teardowns) that can be referenced across the team. They are not for writing individual content pieces, running specific campaigns, or SEO audits without a brand or launch context.
1216

1317
---
1418

15-
*Part of the [Skill Engineer Master](https://www.agentengineermaster.com/skill-engineer) shared skills library.*
19+
*Part of the [Skill Engineer](https://agentengineermaster.com) shared skills library.*

‎operations/README.md‎

Lines changed: 12 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -1,15 +1,19 @@
11
# Operations
22

3-
Skills for operations, project management, and business process teams.
3+
Skills for operators, founders, and chiefs of staff who need AI agents to run structured decision-making, planning, and research optimization workflows.
44

5-
## Skills in this folder
5+
## Skills
66

7-
- SOP and process documentation
8-
- Project planning and sprint management
9-
- Vendor and supplier management
10-
- Workflow automation design
11-
- Business intelligence and reporting
7+
| Skill | Description | When to Use |
8+
|---|---|---|
9+
| [auto-research](auto-research/) | Autonomously optimizes any Claude skill or CLAUDE.md file through a closed hypothesis→test→evaluate→keep/discard loop — produces an updated skill file and an iteration dashboard | Improving a skill's output quality against measurable criteria; running autonomous evals on a skill; setting up a self-improving optimization loop |
10+
| [morning-brief](morning-brief/) | Generates a structured morning brief from current priorities, active task board, latest competitor digest, and content pipeline status — writes it to the daily review file | Daily morning planning and intention setting; automated weekday briefing |
11+
| [stress-test](stress-test/) | Three-phase strategic decision analysis combining verbalized sampling, customizable analytical lenses, and a structured decision brief — surfaces non-obvious tail-distribution insights that mode-collapsed prompting misses | Any high-stakes decision: founder pivots, category selection, brand positioning, workstream prioritisation, pricing, hiring |
12+
13+
## When to use this domain
14+
15+
Use operations skills when the stakes are high enough to warrant a structured process — not a quick answer. These skills enforce methodological rigour: the stress-test runs three phases before producing a recommendation, auto-research runs closed-loop evals before updating a skill. They are the right choice when you want an AI agent to slow down, sample broadly, and surface what standard prompting would miss.
1216

1317
---
1418

15-
*Part of the [Skill Engineer Master](https://www.agentengineermaster.com/skill-engineer) shared skills library.*
19+
*Part of the [Skill Engineer](https://agentengineermaster.com) shared skills library.*

0 commit comments

Comments
 (0)