CtrlK
BlogDocsLog inGet started
Tessl Logo

axir-language-backend

Use when adding or changing generated AxIR language backends in this repo, including target registration, codegen templates, package metadata, examples, conformance, and verification. This is a repo-maintainer skill and must not be emitted into generated Ax packages.

67

Quality

84%

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

SKILL.md
Quality
Evals
Security

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.

A lean, rule-dense body with concrete file paths, commands, and an acceptance bar that doubles as validation — genuinely actionable for a compiler maintainer. The main defects are a dangling `regex.md` reference with no bundle file behind it and the absence of any implementation example to bridge 'inspect the seams' to 'write the code'.

Suggestions

Fix the dangling reference in the Session validation section: either add `references/regex.md` with the UTF-16/pattern-invariant details it promises, or remove/replace the `regex.md` mention with the inline facts actually needed.

Tighten the long compound bullets in Backend Implementation Rules (e.g. the placeholder-body enumeration and the conformance-coverage bullet) into shorter, separate rules so each constraint is scannable.

Add a brief registration example (or a pointer to a concrete existing target) showing what a target registration and emitter stub look like, so the touchpoint list translates directly into a first edit.

DimensionReasoningScore

Conciseness

Dense, rule-list format with no padding and no explanation of concepts Claude already knows — every bullet is repo-specific policy. Not a 5 because several bullets are over-stuffed compound sentences (e.g. "Concrete public generated methods must never be placeholder-only bodies such as `pass`, `return None`, `return nil`, `Value::Null`, empty vectors, or generic 'not implemented'/'unsupported' fallbacks") that could be tightened.

4 / 5

Actionability

Highly concrete guidance: exact file paths (`tools/axir/internal/axir/codegen.go`, `axir_test.go`, `scripts/run-example.mjs`), copy-paste verification commands (`npm run test:axir`, `axir verify --targets python,java,cpp,<new-target>`), and named suites and doc locations. Not a 5 because the implementation path itself is never shown — no snippet or template of what target registration or an emitter actually looks like, leaving minor gaps between 'inspect these seams' and 'do it this way'.

4 / 5

Workflow Clarity

Clear logical sequence via section order (First Checks → Implementation Rules → Required Touchpoints → Acceptance Bar → Avoid) with an explicit Acceptance Bar serving as validation checkpoints (test:axir, axir verify, generated-output audits). Not a 5 because the sequence is unnumbered and there is no explicit fix-and-re-run feedback loop; error recovery is only implied by the Avoid section.

4 / 5

Progressive Disclosure

Sections are well-organized and appropriately sized, but the body references "see `regex.md`" (Session validation section) and no such file exists — no references/, scripts/, or assets/ directories are present in the bundle, so this is a dangling pointer. The reference is also buried inline in prose without a link or dedicated section. Not a 4 because a cited reference file that does not exist is a navigation failure, not a minor organization gap; not a 2 because the overall structure is otherwise sound.

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 description with an explicit when-clause, a comprehensive enumeration of the backend surface, and active disambiguation from adjacent product skills. The only weakness is missing concrete target-language synonyms (Python, Java, C++, Go, Rust) that would improve natural trigger matching.

DimensionReasoningScore

Specificity

The description enumerates the complete backend surface with concrete objects: "adding or changing generated AxIR language backends in this repo, including target registration, codegen templates, package metadata, examples, conformance, and verification" — multiple specific actions with comprehensive coverage of the domain, matching the level-5 anchor. Not a 4 because no coverage gaps remain for this skill's scope.

5 / 5

Completeness

Explicitly answers both: "Use when adding or changing generated AxIR language backends in this repo" is a concrete when-clause, and the enumerated registration/templates/metadata/examples/conformance/verification list plus the boundary statement ("must not be emitted into generated Ax packages") gives a clear what. Not a 4 because the when-clause is fully explicit with concrete trigger phrasing rather than vague.

5 / 5

Trigger Term Quality

Good natural keywords a maintainer would say — "AxIR", "language backends", "codegen", "conformance", "verification", "package metadata" — but common variations like concrete target-language names (Python, Java, C++, Go, Rust) that appear in the body are missing from the description. Not a 5 because a user asking to "add a Rust target" or "new generated language" would not match on explicit synonyms in the description alone.

4 / 5

Distinctiveness Conflict Risk

The tightly-scoped niche ("generated AxIR language backends in this repo") plus explicit disambiguation ("This is a repo-maintainer skill and must not be emitted into generated Ax packages") gives minimal conflict risk with other skills. Not a 4 because it actively distinguishes itself from adjacent skills (Ax product skills) rather than merely being distinct.

5 / 5

Total

19

/

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

referenced_paths_exist

Referenced path issues: 1 missing

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.