Name: ax-go-refine
Rating: 60.8 (1 reviews)
Author: ax-llm

ax-go-refine

Use when writing Go code with `github.com/ax-llm/ax/packages/go` for reward-scored generation, iterative candidate improvement, evaluator feedback, and optimizer-backed refinement patterns.

Quality

76%

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 ./website/static/go/.well-known/agent-skills/ax-go-refine/SKILL.md

Quality

Content

65%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.

The body is exceptionally concise and fact-dense with well-organized sections, but its core code pattern is non-executable pseudocode with inconsistent package prefixes, the workflow is implicit rather than sequenced, and its reference pointers are unlinked and point at files absent from the bundle. These are fixable gaps that keep otherwise strong material from being directly actionable.

Suggestions

Make the Core Pattern executable: add imports, use the consistent `axllm` prefix (matching `axllm.AxGEPA` in the API surface), and show how `reflectionClient`, `request`, and `evaluator` are constructed — ideally copied from a working example in `examples/`.

Turn the Guardrails into an ordered workflow with checkpoints: consult `axir-capabilities.json`/examples for native syntax → draft the call → verify locally with the scripted no-key transport → only then use a real network provider.

Fix reference signaling and existence: either ship `API.md`, `axir-api.json`, `axir-capabilities.json`, and `examples/` in the bundle or remove those pointers, and present them as markdown links in a dedicated 'Reference materials' section rather than a facts bullet.

Dimension	Reasoning	Score
Conciseness	The body is lean bullet-style facts — "Runtime profiles: `javascript-goja`", "Real network support: yes" — with zero padding and no explanation of concepts Claude already knows (no Go tutorials, no library primers). Every section carries non-obvious, package-specific information, matching the 'every token earns its place' anchor.	5 / 5
Actionability	The Core Pattern snippet `engine := ax.NewGEPA(reflectionClient, nil); result := engine.Optimize(request, evaluator)` is effectively pseudocode: `reflectionClient`, `request`, and `evaluator` are undefined, no imports are shown, and the `ax.` prefix contradicts the API surface's `axllm.AxGEPA`. The "Relevant API Surface" is name-only, and the referenced `examples/` are absent from the bundle, so guidance is concrete but incomplete — squarely the 3 anchor, not the 4 anchor's 'minor gaps'.	3 / 5
Workflow Clarity	Guidance is scattered across "When To Use" and "Guardrails" ("Start from package examples... Use `no-key` examples for deterministic local checks") with no explicit sequence or validation checkpoints; the implied check-examples-then-verify-locally-then-network flow is present but unordered and implicit, matching the 3 anchor rather than the 4 anchor's clear sequenced checkpoints.	3 / 5
Progressive Disclosure	The body points to `API.md`, `axir-api.json`, `axir-capabilities.json`, and `examples/`, but only as a bare "Package Facts" bullet list with no markdown links or navigation section, and none of these files exist in the bundle (no references/, scripts/, or assets/ directories). Structure exists and references are named but not clearly signaled and unverifiable — the 3 anchor, not the 4 anchor's 'references mostly clear' with only minor gaps.	3 / 5
	Total	14 / 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, niche-scoped description that explicitly states both what it covers and when to use it, with concrete domain actions and an exact package identifier. Its main weaknesses are a single narrow trigger condition and missing common synonyms (e.g., "golang", "prompt optimization") that would improve natural triggering.

Dimension	Reasoning	Score
Specificity	The description lists several concrete action areas — "reward-scored generation, iterative candidate improvement, evaluator feedback, and optimizer-backed refinement patterns" — which matches the 'several specific actions; minor gaps' anchor rather than the 5 anchor, since 'optimizer-backed refinement patterns' is abstract and coverage of related tasks (e.g., local verification, porting) is thin.	4 / 5
Completeness	Both are answered explicitly: what ("for reward-scored generation, iterative candidate improvement, evaluator feedback, and optimizer-backed refinement patterns") and when ("Use when writing Go code with `github.com/ax-llm/ax/packages/go`"). It falls short of the 5 anchor because the when is a single narrow trigger condition rather than multiple concrete trigger phrases a user would naturally say.	4 / 5
Trigger Term Quality	Natural terms like "Go code", "evaluator feedback", and "optimizer" give good keyword coverage, but common variations such as "golang", "prompt optimization", or "GEPA" are missing, so it does not reach the 5 anchor's comprehensive synonym coverage.	4 / 5
Distinctiveness Conflict Risk	The exact package path `github.com/ax-llm/ax/packages/go` combined with framework-specific terms (reward-scored generation, GEPA-style refinement) carves out a clear niche with minimal conflict risk against other skills.	5 / 5
	Total	17 / 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: website/static/go/.well-known/agent-skills/ax-go-refine/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.