phoenix-cli

Debug LLM applications using the Phoenix CLI. Fetch traces, spans, and sessions, annotate them, analyze errors, inspect datasets, review experiments, query annotation configs, and use the GraphQL API. Use whenever the user works with a Phoenix instance from the terminal.

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

71%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.

Highly actionable CLI reference with strong command coverage and validation around destructive/setup operations, but it is a long monolithic file that could be tightened and split into reference files to improve token efficiency and navigation.

Suggestions

Move the detailed filter-expression sections (span/trace/session DSLs) and the per-resource JSON shape references into separate files under references/, keeping SKILL.md a concise overview with clearly signaled one-level-deep links.

Condense the px setup prose (exit codes, tracesVerified/verification semantics) into a compact table or checklist to reduce token cost while preserving the guidance.

Trim explanatory prose around the command catalogs — the commands are self-explanatory to Claude, so many of the inline clarifying sentences can be dropped or shortened.

Dimension	Reasoning	Score
Conciseness	Mostly efficient concrete commands, but the ~600-line body includes lengthy prose on the filter-expression DSLs and the px setup exit-code/verification details that could be tightened, so it sits between efficient and lean.	3 / 5
Actionability	Extensive copy-paste-ready, executable commands with jq pipelines, specific flags, exit codes, and JSON shapes covering the common cases across every resource.	5 / 5
Workflow Clarity	The px setup flow has explicit verification feedback (tracesVerified, exit codes 6/3/2/0) and destructive deletes are gated by an env var plus confirmation that exits without deleting; most checkpoints are present with only minor gaps in the per-command reference sections.	4 / 5
Progressive Disclosure	Decent section structure, but the skill is a monolithic single file with no bundle files and content that could live in separate references (filter-expression DSLs, JSON shapes) is inlined rather than split one level deep.	3 / 5
	Total	15 / 20 Passed

Description

92%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 that names a comprehensive set of concrete actions and pairs them with an explicit trigger clause. The only minor gap is a few natural synonyms that could broaden trigger matching.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Fetch traces, spans, and sessions, annotate them, analyze errors, inspect datasets, review experiments, query annotation configs, and use the GraphQL API' — giving comprehensive coverage of the CLI's capabilities.	5 / 5
Completeness	Explicitly answers both 'what' (debug LLM apps via the listed actions) and 'when' ('Use whenever the user works with a Phoenix instance from the terminal') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Good natural-term coverage (traces, spans, sessions, annotate, errors, datasets, experiments, GraphQL, Phoenix, terminal), but a few user-plausible synonyms (observability, telemetry, Arize) are absent and no file extensions apply, so it falls just short of the comprehensive anchor.	4 / 5
Distinctiveness Conflict Risk	A clear niche (the Phoenix CLI / Phoenix LLM-observability instance from the terminal) with distinct triggers and minimal overlap with other skills.	5 / 5
	Total	19 / 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
skill_md_line_count	SKILL.md is long (608 lines); consider splitting into references/ and linking	Warning

	Total	15 / 16 Passed

Repository: Arize-ai/phoenix
Path: .agents/skills/phoenix-cli/SKILL.md
Commit: e127482

Reviewed: 6 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.