Skip to content
Limin Zheng

AI Skills & Workflows

New AI Agent Skill

Package a repeated AI behavior into an installable and testable Skill.

Choose this when
Choose this for a new scoped Skill; for an existing Skill choose Skill installation, migration & repair.
Scope unit
One Skill behavior, target host and test matrix
Possible delivery format
Installable ZIP + setup guide + test evidence
Prepare a brief →

You send · You receive

What to send

  • Target AI host and version (e.g., Codex or Claude Code)
  • Successful and failing input/output examples
  • Rules, permissions, tool access and forbidden behavior
  • Confirmation the recipient has a compatible host/environment

Start by describing the materials you already have; missing details can be added when we agree the scope. Before work begins, we will confirm the materials, permissions and working conditions needed for your specific task.

Possible outputs

  • Skill package and scripts needed for the agreed host
  • Installation and use guide with examples
  • Executed test cases and known limitations

Key check

Test normal, missing-information and boundary cases on the agreed host/version.

Verifiable acceptance · 4
  • The client receives a package, setup guide and run records, can install on the agreed host/version with required permissions, and can reproduce agreed case outputs and known limits.
  • Install from a clean directory or defined environment
  • Run agreed normal, missing-information and edge cases
  • Do not claim portability to untested hosts

Hypothetical brief & output structure

Illustration · hypothetical

Illustrative example — not a completed client order

The request

I process a batch of support feedback every week. Build a reusable Codex Skill that turns it into source-linked action items without guessing missing facts.

Sample artifacts

  • feedback-workpack-skill.zipinstallable package for the named host
  • README.mdsetup and repeat-use guide
  • test-results.mdnormal, missing-data and edge-case runs

File names are illustrative

Other tasks of this kind

  • Create a Skill for a named host to organize weekly learning notes with source citations.
Output structure / test path

Workflow: load bounded batch → classify → create source_id-linked tasks → flag missing data and human review

Write a brief from this example →

Scope boundaries

Client needs the target host; ZIP does not include AI access or cross-host compatibility.

Not included by default

  • Providing an AI subscription or API quota
  • Universal host support without testing
  • Ongoing monitoring or enterprise-scale evaluation

What affects the quote

Behavior complexity, host count, tool dependencies and test matrix.

Actual project references

  • Reflection Companion ↗

    Published Agent Skill (Alpha)

    Host support follows the public project documentation; Alpha does not mean production-ready on all platforms.

  • Roleward Job Hunting ↗

    Published Agent Skill (Alpha)

    The Codex local workspace is the fully accepted host as of the catalog review; other hosts are not guaranteed.

  • Score Simplifier ↗

    Published Agent Skill (Alpha)

    Published examples verify score-sourced simplification, engraving and PDF/MusicXML/MIDI checks; they do not prove accurate transcription of arbitrary audio recordings.

Self-directed work and published projects demonstrate relevant practice. They are not paid-client delivery records or guarantees for this task.

Start with a short note

Three answers are enough. I will reply with what else is needed and whether a diagnosis comes first. You get a quote before any work starts.

Prepare a brief →

New AI Agent Skill

Start with a short note

Three answers are enough. I will reply with what else is needed and whether a diagnosis comes first. You get a quote before any work starts.

What to send
  • Target AI host and version (e.g., Codex or Claude Code)
  • Successful and failing input/output examples
  • Rules, permissions, tool access and forbidden behavior
  • Confirmation the recipient has a compatible host/environment
Illustration · hypothetical

Illustrative example — not a completed client order

The request

I process a batch of support feedback every week. Build a reusable Codex Skill that turns it into source-linked action items without guessing missing facts.

Example inputs

  • Three de-identified batches and desired task-table examples
  • An available Codex workspace; read limited input directory; no automatic sending or production writes

To start, it helps to mention

  • What repeated task should the Skill perform, and on which AI host?
  • Give one desired input/output and the failure it must avoid.

Confirm before work begins: Available compatible host/permissions, sample inputs and edge-case tests

Public/de-identified information only. Private access requires separate authorization.

Open email draft →

Preview the text
Services brief — A01 New AI Agent Skill

First contact does not require a complete specification. Unknown fields may be blank or marked unknown.

What do you want to end up with?


What do you already have?


When do you need it? Any budget range?

Select and copy the brief manually if scripting or clipboard access is unavailable.

Public email is a contact starting point. This site has no booking, payment or file-upload service. zhenglimin.design@gmail.com

← Back to category
Prepare a brief →