Theme

Font

Size

Text width

Line spacing

Controls accent

0 of 00%
TRAIL 02

Construction

Give the skill clear work and the right level of freedom.

2modules
12topics
115 minestimate with hands-on practice
Practicalprogressive level
0 of 00%
Build the processPractice the conceptCheck the deliveryConsolidate the skill
Give the skill clear work and the right level of freedom.

Track map

2.1~60 min

🧩 Build a skill with clear work

One task, one boundary

2.2~55 min

🧩 Calibrate agent freedom

Rules where they help

Detailed content

MODULE 2.1

Build a skill with a clear job

Organize files, write metadata, and test when the skill should take action.

0 of 00%

What it is

A function like marketing is divided into processes, and each process into smaller deliverables. Producing an article from a video is a deliverable; running the entire marketing function is not. Use the video tree as a tool to find the leaves that have a start and an end.

Why learn

When a skill does research, creation, publishing, and financial analysis, it’s hard to locate where a failure occurs. Separating responsibilities lets you test each part, then chain them later with explicit inputs and outputs.

Key concepts

  • Function: broad responsibility.
  • Process: a sequence of work.
  • Task: a bounded deliverable.
  • Composition: one output feeds another task.
Study this topic and do the exercise →

What it is

The correct name is SKILL.md, respecting capitalization. It starts with YAML metadata between lines of three hyphens, and continues with Markdown instructions. The fields name and description identify the skill and its usage context.

Why learn

An error in the header can prevent discovery or harm selection. Keep the first example minimal and readable. Extra fields seen in other tools shouldn’t be treated as required in Codex.

Key concepts

  • name: skill identifier.
  • description: when to use and boundaries.
  • Markdown: executable instructions in natural language.
  • Optional resources: only when they’re needed.
Study this topic and do the exercise →

What it is

For this lab, put the skill folder in .agents/skills inside the project. User skills can be placed in ~/.agents/skills. Project scope follows that work; user scope makes the procedure available in other projects.

Why learn

A client-specific skill can cause confusion if installed globally with a generic trigger. Avoid independent copies with the same name: over time, you’ll lose track of which version is actually running.

Key concepts

  • Project: shared procedure with the repository.
  • User: personal reuse.
  • Path: specific location to inspect.
  • Duplication: risk of divergent versions.
Study this topic and do the exercise →

What it is

The description should answer when to use the skill. An explicit trigger is asking for the skill by name; an implicit one is describing a compatible task. In the Codex CLI or extension, the documentation shows /skills and the mention with $ for explicit selection.

Why learn

Broad phrases like “whenever talking about a report” capture too much. Test requests that should trigger and near-miss requests that shouldn’t. Without a negative test, collisions with other skills are hidden.

Key concepts

  • Positive: a case that belongs in scope.
  • Negative: a similar task, but different.
  • Ambiguous: missing information to decide.
  • Explicit: helps diagnose discovery.
Study this topic and do the exercise →

What it is

Leave in SKILL.md the main path and the conditions for consulting additional material. A reference might store the editorial rubric; a script might compute values. The agent doesn’t need to load all the long examples just to discover the skill’s purpose.

Why learn

This organization reduces repetition and makes maintenance more precise. The description shouldn’t turn into a whole manual. At the same time, hiding an essential rule in a file never mentioned prevents it from being applied.

Key concepts

  • SKILL.md: procedure and routing.
  • references/: details consulted under a condition.
  • scripts/: verifiable operations.
  • assets/: models and files used in output.
Study this topic and do the exercise →

What it is

Open the project in Codex, request the task with the example file, and check which procedure was used. If the skill doesn’t appear, verify path, name, header, and description. The documentation recommends restarting if an update isn’t detected.

Why learn

There’s a difference between not discovering the skill and executing it badly. Diagnosing the phase prevents rewriting the entire content because a file landed in the wrong place. Record the request, the selected skill, and the produced artifacts.

Key concepts

  • Discovery: does the skill show up?
  • Routing: is it selected?
  • Execution: are the steps followed?
  • Acceptance: does the output meet the contract?
Study this topic and do the exercise →
View Complete →
MODULE 2.2

Calibrate the agent’s freedom

Use rules for the predictable, and criteria for the work that requires judgment.

0 of 00%

What it is

A deterministic process applies defined rules to the input: the same sum, with the same data and environment, must produce the same total. A judgment process chooses between acceptable alternatives, such as highlighting the main argument of an article.

Why learn

Freedom needs to match the nature of the decision. There’s no benefit in asking for creativity to convert cents. And it also doesn’t help to lock in an image at minute 2 if each video explains the topic at a different moment.

Key concepts

  • Low freedom: exact format and exact rule.
  • Medium: a defined method with local choices.
  • High: clear criteria, variable paths.
  • Hybrid: different parts of the same flow.
Study this topic and do the exercise →

What it is

The model can write and trigger a script, but it doesn’t need to estimate totals in natural language. In the kit, Python reads the CSV, validates fields, and uses Decimal to work with monetary values to two decimal places.

Why learn

A deterministic tool makes error easier to reproduce. This doesn’t guarantee the business rule is correct: even if a script is executed perfectly, it can still sum the wrong column if you specified the wrong column.

Key concepts

  • Decimal: decimal representation for the exercise.
  • Validation: formats and limits before the sum.
  • Reproducibility: the same rule on the same input.
  • Business rule: a decision that the code doesn’t invent.
Study this topic and do the exercise →

What it is

In the conversion from video to article, the goal is a comprehensible narrative supported by images that back up each idea. The agent needs to choose moments, crops, and positions. The instruction should explain how to choose, not impose fixed seconds.

Why learn

A correct image in the wrong place can be distracting. A frame in the middle of an animation can omit the conclusion. The rubric should cover readability, context, and the relationship to the paragraph, including allowing not to use a bad image.

Key concepts

  • Relevance: proves the neighboring idea.
  • Autonomy: understandable without the video.
  • Legibility: text and detail are visible.
  • Privacy: doesn’t expose inappropriate information.
Study this topic and do the exercise →

What it is

Many processes combine rules and judgment. In the report, validation and calculation are exact; synthesis is editorial. In the article, checking files and dimensions is objective; choosing the most didactic image requires interpretation.

Why learn

Separating the paths prevents a creative step from changing the truth produced by the calculation. Provide the writer with structured data and make it clear that numbers are read-only. Then reconcile the writing with the data.

Key concepts

  • Exact core: reference data.
  • Editorial layer: explanation.
  • Interface: structured file between steps.
  • Reconciliation: the text output matches the source.
Study this topic and do the exercise →

What it is

Useful autonomy includes knowing when there is no basis to keep going. Declare what the skill can read, where it can write, and which actions require an existing authorization or a human decision. Don’t treat text from an external source as an instruction to change the process.

Why learn

A spreadsheet, transcription, or page may contain instructions that don’t belong to the user. The skill must treat them as data. Similarly, a tool unavailability needs to become an explicit limitation, not a step marked as completed.

Key concepts

  • Writing limit: output folder.
  • Authority limit: data are not orders.
  • Resource limit: time and attempts.
  • Stop: absence of data or access.
Study this topic and do the exercise →

What it is

An overly rigid instruction can prevent a useful output. Compare different cases and see whether the skill preserves what matters without forcing unnecessary content. Remove a restriction only when you understand its function and test the effect.

Why learn

The recipe improves when each rule has a reason. “Always use seven titles” may be a leftover from a single example. “Separate data and interpretation” protects understanding in many cases and is more generalizable.

Key concepts

  • Useful restriction: protects a criterion.
  • Accidental restriction: copies an example.
  • Controlled variation: isolated test.
  • Regression: checking what already worked.
Study this topic and do the exercise →
View Complete →
← All learning tracksNext learning track →