CtrlK
BlogDocsLog inGet started
Tessl Logo

retail-virtual-tryon

Creates virtual try-on agents supporting image and video (catwalk animation) try-ons on Google Cloud (Gemini image models and Veo on Gemini Enterprise Agent Platform). Handles resource setup, user photo uploading, image/video generation pipelines, local testing, and evaluation.

54

Quality

61%

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 ./plugins/retail/virtual-tryon/SKILL.md

The canonical home for this skill is retail-virtual-tryon in google/adk-samples

SKILL.md
Quality
Evals
Security

Quality

Content

63%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 delivers a highly actionable, well-sequenced workflow with exact commands, question scripts, troubleshooting, and a completion checklist. Its weaknesses are duplication across sections, an undefined "Mode 3" reference, and a bundled architecture reference file that is never surfaced, which hurt conciseness and progressive disclosure.

Suggestions

Link `references/tryon-architecture.md` from the body (e.g. under a "How it works" pointer) so the existing reference file is discoverable.

Fix the "Mode 1, 2, or 3" inconsistency and reconcile the differing `tryon_model` defaults ("flash" vs "gemini-2.5-flash-image") between Mode 1 and Mode 2.

Deduplicate: state the `setup.py` invocation and Q-MODE rules once, and either drop the redundant "Testing & Verification" env-var block or explain when it applies versus design-spec.md.

DimensionReasoningScore

Conciseness

The body is dense and operational with little concept re-teaching, but contains notable duplication: the Q-MODE block is restated under Execution Rules, the `setup.py --config ./design-spec.md` command appears three times, and the "Testing & Verification" env-var block overlaps with the design-spec/config mechanism. Fits anchor 3 ("mostly efficient but could be tightened") rather than 4 due to the repeated sections.

3 / 5

Actionability

Provides copy-paste-ready Q-MODE text, exact commands (`.venv/bin/python "$SKILL_DIR/scripts/setup.py" --config ./design-spec.md`), and runnable smoke tests. Not a 5: the CRITICAL WARNING references a "Mode 1, 2, or 3" selection when only two modes are defined, `tryon_model` defaults differ between modes ("flash" vs "gemini-2.5-flash-image") without explanation, and the smoke-test imports assume `scripts` is importable from the workspace without stating the prerequisite.

4 / 5

Workflow Clarity

Steps are clearly ordered (Q-MODE interview → write design-spec → setup.py → launch sandbox/export), with "Verify each step succeeded before moving to the next", a troubleshooting error→fix table, and a Completion Checklist. Not a 5: some checkpoints are implicit ("Stream the output. On success..." lacks a failure path), the undefined "Mode 3" reference muddies the branching, and Mode-2's deploy steps lack verification of the deployment itself beyond fetching the URL.

4 / 5

Progressive Disclosure

Scored against the actual bundle: `references/tryon-architecture.md` exists but is never mentioned or linked anywhere in the body, so a prepared reference is buried, and long operational blocks (Mode-2 export instructions, sandbox UI walkthrough) are inlined rather than split out. Fits anchor 3 ("references present but not clearly signaled; content that should be separate is inline"); the well-referenced script paths keep it from dropping lower, but the unlinked architecture reference prevents a 4.

3 / 5

Total

14

/

20

Passed

Description

58%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.

The description communicates a specific, technically grounded capability set with concrete actions and a distinct niche on Google Cloud. Its main weakness is the complete absence of a "Use when..." trigger clause and natural user-facing synonyms like "fitting room" or "try on clothes", which limits discoverability and caps completeness.

Suggestions

Append an explicit trigger clause, e.g. "Use when the user wants virtual try-on / fitting-room features for e-commerce, catwalk videos of clothing, or try-on for eyewear and accessories."

Add natural synonyms users would actually say ("try on clothes", "fitting room", "e-commerce product try-on") alongside the existing technical terms.

Clarify the vague terms "resource setup" and "evaluation" with one or two concrete specifics (e.g. "provisions GCS buckets and Gemini APIs; verifies output quality").

DimensionReasoningScore

Specificity

Lists several concrete actions ("Creates virtual try-on agents", "Handles resource setup, user photo uploading, image/video generation pipelines, local testing, and evaluation") with named technologies (Gemini, Veo, Google Cloud). Not a 5 because "resource setup" and "evaluation" are broad and coverage has minor gaps.

4 / 5

Completeness

The "what" is clear and specific (creating try-on agents with image/video modes on Google Cloud), but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3 per the rubric.

3 / 5

Trigger Term Quality

Includes relevant keywords like "virtual try-on", "catwalk animation", "Google Cloud", and "Gemini"/"Veo", but misses natural user phrases and synonyms such as "fitting room", "try on clothes", "e-commerce", or "retail" (the latter appears only in the skill name). Falls between anchor 3 and 4; missing common variations keeps it at 3.

3 / 5

Distinctiveness Conflict Risk

Occupies a clear niche (virtual try-on agents on Google Cloud with Gemini/Veo) that is mostly distinct, with only minor overlap risk against general image-generation or retail skills. Not a 5 because adjacent image-editing/commerce skills could plausibly compete for similar queries.

4 / 5

Total

14

/

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

metadata_field

'metadata' should map string keys to string values

Warning

Total

15

/

16

Passed

Repository
google/adk-samples
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.