create-agent

MUST be used whenever scaffolding a new Atlas agent project. Creates agent.yaml, README.md, and a starter eval/cases.yaml with the correct structure, tool definitions, and instructions. Do NOT manually write agent.yaml from scratch — this skill handles the full scaffold. Triggers: create agent, new agent, scaffold agent, agent project, agent template, cognite agents create.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

75%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, actionable scaffold skill with clear sequencing and built-in safeguards. It could improve with an explicit CLI-failure recovery loop and slightly tighter deduplication between Step 3 and the Guard rails section.

Suggestions

Add a feedback loop in Step 2 for when `npx … agents create` fails (inspect the error, adjust flags, retry) to lift workflow_clarity toward 5.

De-duplicate the 'Do NOT add tools: []' guidance so it lives in one place (either Step 3 or Guard rails), trimming conciseness redundancy.

Consider moving the 14-row tool-type table into a references/ file referenced from the body to strengthen progressive_disclosure for what is otherwise a self-contained skill.

Dimension	Reasoning	Score
Conciseness	Lean tables and code blocks with no over-explanation of concepts Claude already knows; minor redundancy where the Guard rails section repeats Step 3's 'Do NOT add tools: []' rule.	4 / 5
Actionability	Provides the concrete `npx @cognite/cli@latest agents create` command with explicit flag-handling rules, a tool-type table, and a copy-ready YAML example; templated placeholders are appropriate for a scaffold but keep it just short of fully copy-paste-ready.	4 / 5
Workflow Clarity	Steps 1–4 are clearly sequenced with a directory-exists guard and a Step 4 file-confirmation checkpoint; the missing piece is an explicit error-recovery loop for when the CLI command itself fails.	4 / 5
Progressive Disclosure	Self-contained with no bundle files and well-organized sections; the sizeable inline tool-type table is reasonable but represents a minor organization gap versus moving it to a reference file.	4 / 5
	Total	16 / 20 Passed

Description

95%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description that clearly answers both what the skill does and when to invoke it, with a rich set of natural and product-specific trigger terms. The only soft spot is action-verb variety, which is uniformly 'creates'.

Dimension	Reasoning	Score
Specificity	Names concrete outputs ('Creates agent.yaml, README.md, and a starter eval/cases.yaml … tool definitions, and instructions'), but the action verbs are uniform (all 'creates') leaving minor coverage gaps compared to the comprehensive multi-action anchor 5.	4 / 5
Completeness	Explicitly states what ('Creates agent.yaml, README.md … with the correct structure, tool definitions, and instructions … handles the full scaffold') and when ('MUST be used whenever scaffolding a new Atlas agent project' plus a concrete trigger list).	5 / 5
Trigger Term Quality	'Triggers: create agent, new agent, scaffold agent, agent project, agent template, cognite agents create' provides comprehensive natural-term coverage including synonyms and the product-specific phrase.	5 / 5
Distinctiveness Conflict Risk	Clear niche (Atlas/Cognite agent scaffolding) with distinct product-specific triggers ('cognite agents create', 'Atlas agent project'), giving minimal conflict risk with other skills.	5 / 5
	Total	19 / 20 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
metadata_version	'metadata.version' is missing	Warning

	Total	14 / 16 Passed

Repository: cognitedata/builder-skills
Path: skills/create-agent/SKILL.md
Commit: c87160a

Reviewed: 5 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.