agentsociety-create-agent

Use when creating or revising a custom agent type, when an experiment needs an agent class that does not yet exist in the workspace, or when the agent design must be sized against a simulation budget.

Quality

80%

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

Fix and improve this skill with Tessl

tessl review fix ./extension/skills/agentsociety-create-agent/v1.0.0/SKILL.md

Quality

Content

81%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 dense, well-structured skill body with an explicit validated workflow and strong reference split. Its main weaknesses are references to several non-bundled helper directories and a slightly long Custom Skills explainer.

Suggestions

Either bundle the referenced stages/, checklists/, artifacts/, and subagent-prompts/ files or inline the essential bits so navigation does not dead-end.

Tighten the 'Custom Skills' frontmatter section — the SKILL.md layout example duplicates concepts Claude already knows about YAML frontmatter.

Add a minimal inline code template (or confirm artifacts/templates.md exists) so the generate stage is fully executable without a missing-file hop.

Dimension	Reasoning	Score
Conciseness	Mostly lean and actionable with no basic-concept padding, but the 'Custom Skills' frontmatter explainer and the 'Rewrite from scratch' rationale paragraph are minor instances that could be trimmed; sits at 4 rather than the fully-lean 5.	4 / 5
Actionability	Gives concrete validator commands, file paths, a required-method list, and a Common Mistakes table with fixes; the core generate stage points to 'artifacts/templates.md' which is absent from the bundle, a minor executable gap keeping it at 4.	4 / 5
Workflow Clarity	Clear intake→design→generate→validate→register sequence with a diagram, an explicit validation checkpoint ('Always run … as the final step'), a reviewer→fix→revalidate feedback loop, and referenced checklists; matches anchor 5.	5 / 5
Progressive Disclosure	Good one-level-deep signaling to the bundled references/ files and scripts/validate.py, but the body also points at stages/, checklists/, artifacts/, and subagent-prompts/ paths that are not in the bundle — a real navigation gap keeping it below 5.	4 / 5
	Total	17 / 20 Passed

Description

78%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, third-person description with an explicit 'Use when' trigger and concrete domain phrases. It clearly answers what and when, with minor room to add synonyms and sharpen the scan-modules boundary.

Dimension	Reasoning	Score
Specificity	Quotes 'creating or revising a custom agent type' and 'sized against a simulation budget' — names the domain plus 1-2 concrete actions but is not comprehensive; matches anchor 3, not 4 which requires several specific actions.	3 / 5
Completeness	Explicitly answers both — 'what' via 'creating or revising a custom agent type' and 'when' via three concrete 'Use when…' / 'when…' trigger clauses; matches the anchor-5 example structure.	5 / 5
Trigger Term Quality	Natural domain phrases like 'creating or revising a custom agent type', 'experiment needs an agent class', and 'simulation budget' give good coverage a user would actually say; misses synonyms/extensions so it sits at 4 rather than 5.	4 / 5
Distinctiveness Conflict Risk	Niche triggers ('agent class that does not yet exist in the workspace', 'simulation budget') keep conflict risk low; only weakly disambiguates from scan-modules in the description itself, so it does not reach 5.	4 / 5
	Total	16 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: tsinghua-fib-lab/AgentSociety
Path: extension/skills/agentsociety-create-agent/v1.0.0/SKILL.md
Commit: 13e28b5

Reviewed: 1 day 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.