ax-java-gepa

Use when writing Java code with `dev.axllm:ax` for GEPA, Pareto tradeoffs, reflection clients, metric budgets, optimizer state, and artifacts.

Quality

64%

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 ./packages/java/skills/ax-java-gepa/SKILL.md

Quality

Content

53%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 lean, well-sectioned overview with genuinely useful package facts and guardrails, but it functionally fails as a self-contained guide: the core code pattern is not copy-paste runnable, no workflow checkpoints are given, and every referenced bundle file (API docs, capability manifest, examples) is missing from the skill directory. The skill's own advice — start from package examples — cannot be followed.

Suggestions

Bundle the referenced files (API.md, axir-api.json, axir-capabilities.json, examples/) in the skill directory, or correct the paths so references resolve instead of dangling.

Expand the Core Pattern into one complete runnable example that constructs the reflection client, request, and evaluator, so the snippet is copy-paste ready.

Add a short ordered workflow with an explicit checkpoint, e.g., run optimizer → verify metric budget consumed / inspect Pareto front artifact → handle failed candidates, to give validation guidance.

Dimension	Reasoning	Score
Conciseness	The body is lean — terse fact lists ('Real network support: yes', runtime profiles), a minimal code snippet, and guardrails that each earn their place. Minor trims exist: 'Language: Java.' duplicates what the package coordinate already implies, and the opening sentence restates the description; this matches anchor 4 (efficient, minor over-explanation) rather than anchor 5's 'every token earns its place'.	4 / 5
Actionability	The Core Pattern ('AxGEPA engine = new AxGEPA(reflectionClient, java.util.Map.of()); var result = engine.optimize(request, evaluator);') shows real syntax but is not executable — `reflectionClient`, `request`, and `evaluator` are undefined, and the instruction to 'Start from package examples' points to an examples/ directory that does not exist in the bundle. This matches anchor 3: some concrete guidance but missing key details.	3 / 5
Workflow Clarity	A minimal sequence exists (construct engine → optimize, and 'Use BootstrapFewShot before GEPA' establishes ordering), but there are no validation checkpoints or error-recovery steps for an optimization run (e.g., how to check budgets or inspect a failed candidate). This matches anchor 3 — sequence present, checkpoints missing — and not anchor 4, which requires most checkpoints present.	3 / 5
Progressive Disclosure	The body is structured as an overview pointing to detailed materials ('API.md', 'axir-api.json', 'axir-capabilities.json', 'examples/'), but the bundle contains only SKILL.md — every referenced path dangles, so navigation dead-ends and the disclosure structure is ineffective. Scored against the actual bundle structure per the rubric guideline, this lands at anchor 2; the skill is under 50 lines but it demonstrably needs the external references it claims, so the simple-skill exception does not apply.	2 / 5
	Total	12 / 20 Passed

Description

60%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 highly distinct, well-triggered description for a niche Java optimization package, but it is a trigger-only sentence: it names the domain and its topics without stating what the skill actually does. Adding an explicit action-verb 'what' clause would move it to the top anchors.

Suggestions

Lead with a concrete 'what' statement using action verbs before the trigger clause, e.g., 'Runs the generated GEPA optimizer, seeds demonstrations with BootstrapFewShot, tracks metric budgets, and inspects Pareto fronts and artifacts in Java.

Add 1-2 natural trigger variations users might say, such as 'prompt optimization' or 'Ax package', to broaden keyword coverage.

State both parts explicitly in the pattern 'Does X, Y, Z. Use when...' to fully answer 'what' and 'when'.

Dimension	Reasoning	Score
Specificity	The description names the domain ('writing Java code with `dev.axllm:ax`') and enumerates specific topic objects ('GEPA, Pareto tradeoffs, reflection clients, metric budgets, optimizer state, and artifacts') but contains no concrete action verbs describing what the skill does (no 'runs', 'inspects', 'tracks'). It matches anchor 2 — domain named, actions minimal — rather than anchor 3, which requires 1-2 concrete actions.	2 / 5
Completeness	The 'when' is explicit and specific ('Use when writing Java code with `dev.axllm:ax` for...'), but the 'what' is only weakly implied — the entire sentence is a trigger clause with no separate statement of what the skill does (e.g., 'Runs the generated GEPA optimizer and inspects artifacts'). This sits between anchor 3 (what present, when weak) and anchor 4 (both present); the missing explicit 'what' statement pulls it to 3.	3 / 5
Trigger Term Quality	Terms like 'Java code', 'GEPA', 'Pareto tradeoffs', 'optimizer state', and 'metric budgets' are exactly what a user of this package would naturally say. Coverage is good but missing common variations such as 'prompt optimization', 'Ax package', or file extensions (e.g., '.java'), keeping it just below the comprehensive anchor 5.	4 / 5
Distinctiveness Conflict Risk	The combination of the exact package coordinate `dev.axllm:ax`, 'GEPA', 'Pareto tradeoffs', and 'reflection clients' carves out a clear niche with distinct triggers and minimal risk of firing for an unrelated skill, matching anchor 5.	5 / 5
	Total	14 / 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: ax-llm/ax
Path: packages/java/skills/ax-java-gepa/SKILL.md
Commit: 3f630a0

Reviewed: 3 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.