Domains
Forms
Mechanism
Breaks down how the skill works and what it produces, so you can quickly judge whether it fits your scenario and value.
When you need to evaluate and improve an AI agent's quality using the Agent Platform Eval Service via agents-cli, especially after scaffolding a project or writing eval datasets.
agents-cli eval subcommands.A structured eval loop producing traces, graded results (JSON/HTML), failure analysis, and an optimized agent through 5–10+ iterations per case.
Use Cases
A developer scaffolded a project with /google-agents-cli-scaffold and wants to verify agent quality. They run `agents-cli eval run` to generate traces and grade them, then inspect the HTML report. Using the failure-to-fix table, they adjust prompts and rerun, confirming improvements via `eval compare`. This catches goal-completion and safety issues early.
A team lacks multi-turn conversation data for eval. They use `agents-cli eval dataset synthesize` to simulate users and produce traces, then `eval grade` and `eval analyze` for LLM-based failure clustering. This scales eval coverage without manual dataset writing and reveals categorized failure modes for targeted fixes.
Skill Relationships
Dependency relationships read as "the upper tier points to the lower tier." The current Skill sits in the middle tier — above are Skills that depend on it, below are Skills it depends on.
Tier 1 · These Skills Use Me
Tier 2 · Current Skill
Tier 3 · I Use These Skills
Browse this skill's relationships within its skillset. Click a node to switch the side panel; use the search box to jump to any skill.
Skill File