CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

axiom-vision

Use when implementing ANY computer vision feature — image analysis, pose detection, person segmentation, subject lifting, text recognition, barcode scanning.

Quality

69%

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 ./axiom-codex/skills/axiom-vision/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

60%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 well-organized router/index skill: clear navigation, one-level-deep references, useful anti-rationalization guidance, and good cross-skill boundary handling. Its weaknesses are redundancy (the same routing repeated in four formats) and that all actionable content lives in referenced files that are not present in this bundle, leaving the body itself with pointer-only guidance.

Suggestions

Consolidate the routing into a single canonical form — keep the Quick Reference table and the numbered decision list, and drop the dot digraph and/or the Example Invocations section, which duplicate them.

Fix reference paths so they resolve within the skill bundle (e.g., references/vision-framework.md) or verify the sibling skills/ files exist, since none of the four referenced files are present here.

Inline a minimal copy-paste snippet (e.g., the request/handler skeleton or background-queue dispatch) so the most common case is actionable without opening a referenced file.

DimensionReasoningScore

Conciseness

The same routing information is presented four times (Quick Reference table, dot digraph, numbered decision list, and Example Invocations), which is redundant padding — e.g., the dot graph and the numbered list encode identical routing. Not 2 because there is no over-explanation of concepts Claude already knows and the Anti-Rationalization table adds real value; not 4 because the fourfold repetition is a clear tightening opportunity.

3 / 5

Actionability

Routing is specific per task (each symptom maps to a named file and section), and the Anti-Rationalization table cites concrete details ("IntentValueQuery and SemanticContentDescriptor", "15 min to add background queue"). However, every action is a pointer ("See skills/vision-framework.md") with no executable code or commands in the body, and the referenced files are not present in this bundle — anchor 3 (some concrete guidance but incomplete).

3 / 5

Workflow Clarity

The decision tree plus numbered list gives an unambiguous routing sequence, including the cross-skill redirect to axiom-media for identity clustering. Not 5 because it is a multi-entry router (not a single-action skill) and a few table rows (e.g., document scanning, DataScannerViewController) are only implicitly covered by the tree's "implement feature" branch.

4 / 5

Progressive Disclosure

Clear overview with well-signaled, one-level-deep references organized in a symptom-to-file table — structurally close to anchor 5. Not 5 because the referenced paths (skills/vision-framework.md, skills/vision-ref.md, skills/vision-diag.md, skills/media-intelligence.md) do not exist under this skill's directory (no references/, scripts/, or skills/ bundle is present), so the navigation cannot be verified and the paths point outside the bundle.

4 / 5

Total

14

/

20

Passed

Description

62%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 trigger-focused description with a strong 'Use when' clause and several concrete capability keywords, but it never says what the skill actually offers and omits platform-specific terms (Vision framework, Apple, OCR, QR) that users commonly say. This weakens both completeness and distinctiveness relative to the top anchors.

Suggestions

State the 'what' explicitly, e.g., "Provides implementation patterns, API reference, and diagnostics for Apple's Vision framework — image analysis, pose detection, ...".

Add natural synonyms and platform cues users would say: "OCR", "QR code scanning", "face detection", "iOS/macOS Vision framework".

Narrow the scope claim from "ANY computer vision feature" to the Vision framework specifically to reduce overlap with generic computer-vision skills.

DimensionReasoningScore

Specificity

Lists several concrete capabilities ("pose detection, person segmentation, subject lifting, text recognition, barcode scanning") but "image analysis" is generic and no verb states what the skill itself does. Matches anchor 4 (several specific actions, minor gaps) rather than 5, which requires comprehensive coverage of the actions the skill performs.

4 / 5

Completeness

The 'when' is explicit and strong ("Use when implementing ANY computer vision feature"), but the 'what' is only implied — the description never states what the skill provides (implementation patterns, API reference, diagnostics routing). Falls between anchor 2 ('when' only) and anchor 4 (both present with 'what' stated).

3 / 5

Trigger Term Quality

Good natural keyword coverage ("pose detection", "person segmentation", "text recognition", "barcode scanning") that users would plausibly say. Not 5 because common variations are missing: "OCR", "QR code", "face detection", "Vision framework", or any platform term.

4 / 5

Distinctiveness Conflict Risk

"ANY computer vision feature" names a domain but not the platform, so generic CV skills (e.g., Python/OpenCV) would also match this trigger. Anchor 3 ("somewhat specific but could still overlap with similar skills") fits; it is not 4 because the Vision framework / Apple-platform scoping that would make it distinct is absent.

3 / 5

Total

14

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

CharlesWiltgen/Axiom

Path

axiom-codex/skills/axiom-vision/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.