CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

product-feed-optimizer

Use when the user asks to "optimize my Shopping feed", "fix product disapprovals", "improve product titles/attributes", or "build feed-driven PMax asset groups"; audits and rewrites the Shopping/Performance Max product feed — title/description patterns, required and recommended attributes, GTIN/availability/price hygiene, disapproval triage, and feed-driven asset-group / listing-group structure — informing the ROAS O (Offer) dimension. Not for text ad copy — use ad-creative-builder; not for scoring the account or the RQS — use ad-account-auditor. 商品Feed优化/购物广告Feed/商品标题优化/商品禁投修复

Quality

84%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

73%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-structured instruction skill: an explicit contract, a validated 8-step workflow with a quality-bar checklist, and a real, well-organized one-level-deep reference file. The main costs are repetition of scope/ROAS framing across four sections and a tangle of out-of-bundle cross-references that dilute the lean core.

Suggestions

State the scope boundaries (not ad-creative-builder, not ad-account-auditor, not landing-optimizer) once — in the Scope guard — and drop the repeats from the intro, Skill Contract, and Next Best Skill to reclaim tokens.

Replace step 3's inline attribute enumeration with a pointer to the attribute-priority checklist in references/feed-title-patterns.md, keeping only the required-attribute names inline, to remove duplication.

Consolidate the out-of-bundle ../../../ links (skill-contract, CONNECTORS, SECURITY, humanizer-slop, roas-benchmark) into the Reference Materials section so the instruction steps stay focused on the feed work itself.

DimensionReasoningScore

Conciseness

The body is domain-specific with no basic-concept padding, but the ROAS O(Offer) framing and the 'not ad-creative-builder / not ad-account-auditor' scope disclaimers each repeat three to four times (intro, Skill Contract, Scope guard, Reference Materials, Next Best Skill). That is more than 'minor instances of over-explanation', fitting 'mostly efficient but could be tightened'.

3 / 5

Actionability

Steps 1–8 give concrete executable guidance — named required/recommended attributes, hygiene rules keyed to landing-page truth, an explicit save path, and a NEEDS_INPUT path. Not 5 because the worked patterns (per-platform character limits, title formulas, before/after examples) live only in the reference file with no inline quick example.

4 / 5

Workflow Clarity

Eight clearly sequenced steps with ordering rationale ('Triage disapprovals first... a rewritten title on a disapproved item still does not serve'), validation embedded in steps 5–6 (price/availability reconciliation, claim pre-check), an explicit five-point Quality-bar checklist with a fix-or-report feedback loop ('If any item fails, fix it or report it in the handoff — do not ship silently'), and 'Done when' criteria. This matches the anchor with explicit validation, error recovery, and a checklist.

5 / 5

Progressive Disclosure

The bundle reference (references/feed-title-patterns.md) exists, is one level deep, and is clearly signaled both in step 4 and in the Reference Materials section with a description. Not 5 because step 3's required/recommended attribute list duplicates the reference's checklist inline, and the body is threaded with many out-of-bundle ../../../ links (skill-contract.md, CONNECTORS.md, SECURITY.md, humanizer-slop.md, other skills) that add navigation load without being part of the bundle.

4 / 5

Total

16

/

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.

An exemplary description: explicit 'Use when' triggers with natural quoted phrases, a comprehensive and concrete capability list in third person, and explicit disambiguation from adjacent skills. The only gap is the absence of file-format terms (TSV/CSV/XML, Merchant Center export) that users would naturally mention when invoking it.

Suggestions

Add the natural input-format terms users say when asking for this — e.g. 'product feed export (TSV/CSV/XML)' or 'Merchant Center disapprovals' — to the trigger clause to lift trigger-term coverage to comprehensive.

The trailing Chinese keyword string ('商品Feed优化/购物广告Feed/...') duplicates the summary field's role; folding those terms into the trigger clause instead of appending them as a bare keyword list would read more naturally.

DimensionReasoningScore

Specificity

Lists multiple specific concrete actions — 'audits and rewrites the Shopping/Performance Max product feed — title/description patterns, required and recommended attributes, GTIN/availability/price hygiene, disapproval triage, and feed-driven asset-group / listing-group structure' — with comprehensive coverage of the domain. It goes beyond the score-4 anchor ('several specific actions; minor gaps') because every major sub-task of feed optimization is named.

5 / 5

Completeness

Explicitly answers both: what ('audits and rewrites the Shopping/Performance Max product feed...') and when ('Use when the user asks to...') with concrete trigger phrases. It also adds negative scope ('Not for text ad copy — use ad-creative-builder'), which strengthens rather than pads the answer.

5 / 5

Trigger Term Quality

Strong natural quoted phrases users would actually say ('optimize my Shopping feed', 'fix product disapprovals', 'improve product titles/attributes', 'build feed-driven PMax asset groups') plus bilingual synonyms. Not 5 because common format/extension variants users would mention (feed export, TSV/CSV/XML, Merchant Center) are absent; not 3 because coverage clearly exceeds 'some relevant keywords'.

4 / 5

Distinctiveness Conflict Risk

Clear niche (product feed data quality for Shopping/PMax) with distinct triggers and explicit redirects to neighboring skills ('Not for text ad copy — use ad-creative-builder; not for scoring the account or the RQS — use ad-account-auditor'). Minimal conflict risk.

5 / 5

Total

19

/

20

Passed

Validation

81%

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

Validation — 13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_field

'metadata' should map string keys to string values

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

relative_links

Relative link issues: 20 suspicious

Warning

Total

13

/

16

Passed

Repository

aaron-he-zhu/aaron-marketing-skills

Path

ad/research/product-feed-optimizer/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.