CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

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.

Quality

79%

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 ./packages/java/skills/ax-java-agent-optimize/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

72%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-organized body that assumes competence and wastes no tokens, but it stops short of being directly executable: the sole code snippet is an illustrative shape with undefined variables and no complete example is inlined. It also lacks an explicit step sequence for an optimization run, leaving validation checkpoints implicit in the guardrails.

Suggestions

Inline one complete, runnable no-key example (with imports and constructed inputs) in place of or alongside the placeholder Core Pattern snippet, so the guidance is copy-paste ready.

Add a short numbered workflow for an optimization run (construct evaluator -> run engine within explicit budgets -> apply the playbook verification gate -> persist and compare artifacts against baselineScore), turning the implicit guardrail checkpoints into explicit steps.

Expand the API surface section with one-line usage notes or micro-examples for the key entry points (Ax.optimize, AxBootstrapFewShot, OptimizerEvaluator) so the agent knows how each is invoked, not just that it exists.

DimensionReasoningScore

Conciseness

Terse fact bullets ('Language: Java.', 'Real network support: yes.') and dense guardrails with no padding; it assumes Claude's competence by not explaining what GEPA or BootstrapFewShot are. Every line carries non-obvious, package-specific information, matching the lean anchor.

5 / 5

Actionability

The Core Pattern snippet ('new AxGEPA(reflectionClient, java.util.Map.of()); engine.optimize(request, evaluator)') uses undefined placeholders, making it a call-shape sketch rather than executable code, and guardrails defer to 'examples/' instead of inlining a complete example. Not 4 because the gaps (undefined variables, no runnable no-key example) are more than minor; not 2 because the API surface listing and snippet still give real concrete guidance.

3 / 5

Workflow Clarity

No sequenced workflow for an optimization run (build evaluator, run within budgets, apply verification gate, persist artifacts); checkpoints exist only implicitly in guardrails like 'keep only playbook proposals that pass the verification gate' and 'Without new demos, reuse the baseline evaluation'. Not 4 because no explicit sequence or checkpoint steps are listed; not 2 because validation guidance is present, not absent.

3 / 5

Progressive Disclosure

The body is under 50 lines with no bundle files needed, and content is well-organized into clearly headed sections (When To Use, Package Facts, Core Pattern, API Surface, Guardrails) with one-level pointers to package resources (API.md, examples/). Per the rubric's simple-skill note, well-organized sections suffice for a 5.

5 / 5

Total

16

/

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, distinct description with an explicit trigger clause and specific package-anchored terminology. Its main limitation is that capabilities are listed as topic nouns inside the when-clause rather than as explicit action statements.

DimensionReasoningScore

Specificity

Names the exact package ('dev.axllm:ax') and enumerates concrete capability areas ('agent optimization, verified agent-playbook evolution, evaluators, judges, optimizer artifacts, BootstrapFewShot, and GEPA'). Not 5 because these are topic nouns rather than stated actions (e.g., 'create evaluators', 'run optimizers'); not 3 because coverage goes well beyond 1-2 items.

4 / 5

Completeness

The explicit 'Use when writing Java code with dev.axllm:ax for...' clause answers 'when' with concrete triggers, and the 'what' is conveyed by the enumerated feature areas. Not 5 because the 'what' is folded into the when-clause rather than stated as its own explicit action sentence; not 3 because the when is explicit, not merely implied.

4 / 5

Trigger Term Quality

'Java', 'agent optimization', 'evaluators', 'judges', 'BootstrapFewShot', and 'GEPA' are the natural terms a user of this package would say. Not 5 because common variations like 'AxAgent', 'few-shot demos', or 'prompt optimization' are missing; not 3 because keyword coverage is good rather than partial.

4 / 5

Distinctiveness Conflict Risk

Scoped to a specific package ('dev.axllm:ax'), language (Java), and named optimizers (BootstrapFewShot, GEPA), giving it a clear niche with distinct triggers and minimal conflict risk with other skills.

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

Path

packages/java/skills/ax-java-agent-optimize/SKILL.md

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.