CtrlK
BlogDocsLog inGet started
Tessl Logo

ax-java-agent-optimize

Use when writing Java code with `dev.axllm:ax` for agent optimization, verified agent-playbook evolution, evaluators, judges, optimizer artifacts, BootstrapFewShot, and GEPA.

62

Quality

78%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./website/static/java/.well-known/agent-skills/ax-java-agent-optimize/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

68%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.

An exemplary lean, fact-dense overview that assumes Claude's competence and points to package-internal detail instead of inlining it. It falls short on actionability and workflow clarity: the single code snippet is not executable end-to-end and the optimize-mine-verify-evolve loop is never laid out as an explicit sequence with validation checkpoints.

Suggestions

Add one complete, runnable example (imports, client setup, an OptimizerEvaluator callback implementation, engine.optimize call, and artifact persistence) so the Core Pattern is copy-paste ready instead of a call shape with undefined variables.

Lay out the optimization workflow as an explicit numbered sequence with validation checkpoints, e.g., 1) start from a package example, 2) define evaluator, 3) run engine within explicit budgets, 4) only keep playbook proposals that pass the verification gate, 5) persist optimizer artifacts.

Point to specific example files under 'examples/' (or name one per optimizer) instead of the bare directory, and clarify where 'API.md'/'axir-api.json' live relative to the package so navigation does not depend on guessing the package layout.

DimensionReasoningScore

Conciseness

The body is lean and information-dense: every bullet is a non-obvious fact (runtime profiles, no-key transport, TypeScript-API restriction, AxIR-as-source-of-truth rule) with zero re-explanation of concepts Claude already knows. Every token earns its place, matching anchor 5.

5 / 5

Actionability

The Core Pattern gives a real call shape ('AxGEPA engine = new AxGEPA(reflectionClient, java.util.Map.of()); var result = engine.optimize(request, evaluator);') but with undefined variables, no imports, and no complete example for the other advertised tasks (evaluator callbacks, artifact persistence). This is concrete-but-incomplete guidance, matching anchor 3 rather than the minor gaps of anchor 4.

3 / 5

Workflow Clarity

The optimization workflow (mine weaknesses, verification gate, bounded budgets) is implied through 'When To Use' bullets and guardrails rather than sequenced steps, and the verification gate is named but never operationalized. The sequence and checkpoints remain implicit, matching anchor 3; it is above anchor 2 because the guardrails do impose meaningful ordering constraints.

3 / 5

Progressive Disclosure

Well-organized sections with clearly signaled one-level-deep pointers to package detail ('API.md and axir-api.json', 'axir-capabilities.json', 'examples/') plus an appropriately curated inline API surface list. Scored against the actual bundle (only SKILL.md is present), the referenced paths cannot be verified and 'examples/' is a generic pointer, which are the minor organization gaps of anchor 4 rather than the fully verifiable split of anchor 5.

4 / 5

Total

15

/

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 description with an explicit 'Use when' trigger, a clearly named niche package, and a dense list of concrete capability keywords. Its main weakness is that what and when are fused into one clause and the capability list is noun-heavy rather than action-verb-driven.

Suggestions

Split into two sentences: a verb-led 'what' statement (e.g., 'Write and optimize Java agents with dev.axllm:ax: create evaluators, run BootstrapFewShot and GEPA, persist optimizer artifacts') followed by an explicit 'when' clause enumerating trigger phrases users would say.

Add the most likely user-voiced terms such as 'AxAgent' or 'Ax' so the skill triggers on natural requests that omit the full package coordinate.

DimensionReasoningScore

Specificity

The description lists several specific capabilities ('agent optimization, verified agent-playbook evolution, evaluators, judges, optimizer artifacts, BootstrapFewShot, and GEPA') anchored to a named package. It falls short of anchor 5 because these are mostly nouns rather than concrete actions, and above anchor 3 because coverage goes well beyond 1-2 items.

4 / 5

Completeness

Both what ('writing Java code ... for agent optimization, evaluators, optimizer artifacts') and when ('Use when writing Java code with dev.axllm:ax') are explicitly present. However, they are merged into a single clause rather than distinct enumerated trigger phrases, so the when is less explicit than anchor 5 while clearly stronger than anchor 3's missing when.

4 / 5

Trigger Term Quality

Good keyword coverage including 'Java', 'agent optimization', 'evaluators', 'judges', 'BootstrapFewShot', 'GEPA', and the package coordinate 'dev.axllm:ax'. A few natural terms users might say are missing (e.g., 'AxAgent', 'Ax', 'prompt optimization'), keeping it below the comprehensive synonym coverage of anchor 5.

4 / 5

Distinctiveness Conflict Risk

Clear niche with distinct triggers: the 'dev.axllm:ax' package coordinate plus 'BootstrapFewShot', 'GEPA', and 'optimizer artifacts' are highly distinctive. Even the broad 'writing Java code' opening is qualified by the package name, so overlap with generic Java skills is minimal.

5 / 5

Total

17

/

20

Passed

Validation

93%

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

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
ax-llm/ax
Reviewed

Table of Contents

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.