Contribute | AI Agent Library

Contribution principle:

A useful contribution should be specific, public-safe, structured, reusable, and easy to review. Include the intended path and explain how the asset helps builders or operators.

Agents

Submit an Agent

Add a productized assistant with a buyer, pain solved, workflow, outputs, guardrails, and monetization path.

Agent name.
Who it is for.
Main pain it solves.
Inputs and outputs.
Restricted claims.
Suggested path.

Skills

Submit a Skill

Add a reusable capability that can power multiple agents, pages, workflows, schemas, platform packs, or examples.

Capability name.
Required inputs.
Standard outputs.
Workflow steps.
Guardrails.
Example usage.

Prompts

Submit a Prompt Pack

Add a repeatable prompt chain for planning, building, reviewing, launching, auditing, or maintaining AI assets.

Prompt chain name.
Use case.
Prompt sequence.
Inputs required.
Expected outputs.
Review checklist.

Schemas

Submit a Schema

Add JSON schemas that make the repo easier to validate, automate, search, and convert into platform-ready assets.

Schema purpose.
Required fields.
Allowed values.
Example object.
Validation notes.
Related pages or folders.

Examples

Submit an Example

Add a synthetic scenario that shows how an agent, skill, prompt, schema, or workflow should behave.

Scenario.
Inputs.
Missing inputs.
Expected output.
Review notes.
Reusable asset.

Pages

Submit a Static Page

Add or improve HTML pages that make the library more discoverable, trustworthy, useful, or conversion-ready.

Page purpose.
Target user.
Navigation placement.
Required data files.
CTA or next step.
Related issue.

Universal submission checklist

The contribution has a clear purpose and audience. Required inputs and outputs are defined. The suggested repo path is included. The content does not include private client data, secrets, API keys, or protected source material. Funding, business-credit, finance, legal, tax, or platform-related claims are public-safe and non-promissory. Examples are synthetic unless explicitly approved and documented. The contribution can be reviewed without needing hidden context. Related files, folders, or pages are listed.

Issue template

## Contribution Type
Agent / Skill / Prompt Pack / Schema / Example / Page / Portfolio Asset

## Purpose
What does this add or improve?

## Audience
Who is this for?

## Suggested Path
repo/path/here

## Required Inputs
- Input 1
- Input 2

## Expected Output
- Output 1
- Output 2

## Guardrails
- No private data
- No unsupported claims
- Human review where needed

## Acceptance Criteria
- [ ] File added or updated
- [ ] Links checked
- [ ] Public-safe language reviewed
- [ ] Related indexes updated

PR checklist

## Summary
- What changed

## Files Changed
- path/to/file

## Validation
- [ ] Static page opens
- [ ] Navigation works
- [ ] Data files load
- [ ] No secrets or private data
- [ ] Public-safe language reviewed

## Follow-up
- Related issues
- Known limitations

Contribution paths

Contribution	Preferred path	Review focus
Agent spec	`agents/[agent-name].md`	Buyer, pain solved, workflow, outputs, monetization, guardrails.
Skill	`skills/[skill-name]/SKILL.md`	Reusable capability, inputs, outputs, examples, safety notes.
Knowledge base	`knowledge-base/[topic]/`	Source boundaries, public-safe rules, examples, output formats.
Schema	`schemas/[schema-name].schema.json`	Required fields, allowed values, validation utility, examples.
Example	`examples/[category]/example-001.md`	Synthetic inputs, expected output, review notes, guardrails.
Static page	`[page-name].html`	Navigation, style consistency, content quality, links, public safety.