CtrlK
BlogDocsLog inGet started
Tessl Logo

ai-sdk

Vercel AI SDK expert guidance. Use when building AI-powered features — chat interfaces, text generation, structured output, tool calling, agents, MCP integration, streaming, embeddings, reranking, image generation, or working with any LLM provider.

67

Quality

82%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Medium

Suggest reviewing before use

SKILL.md
Quality
Evals
Security

Quality

Content

65%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The content is highly actionable with thorough executable examples, but it suffers from verbosity and redundancy across ~730 lines and weak progressive disclosure — its one reference file is unused from the body and migration/provider material is kept inline. Workflow sequencing is present but lacks explicit validation checkpoints for fragile operations.

Suggestions

De-duplicate the body: the v6 migration pitfalls appear both in 'v6 Migration Pitfalls (Read First)' and again (expanded) in 'Migration from AI SDK 5'; consolidate to one location and link from the other.

Link references/providers.md from the body (e.g., in the Global Provider System and Embeddings sections) and move the detailed provider-routing/migration tables into reference files so the main body is a lean overview.

Add explicit validation/verification checkpoints to multi-step flows — e.g., assert result.files is non-empty after image generation, and confirm gateway routing before relying on it — so workflow clarity reaches the validation-driven anchor.

DimensionReasoningScore

Conciseness

The body is mostly efficient and packed with valuable v6-specific facts, but ~730 lines with notable redundancy (migration pitfalls restated in the Migration section, image-model guidance and 'always use AI Elements' repeated several times) means it could be tightened. It does not earn the 'lean; every token earns its place' anchor, and version numbers are scattered rather than confined to a deprecated section.

2 / 3

Actionability

Provides extensive executable, copy-paste-ready code and exact commands across every capability (generateText, Output.object, tool/needsApproval, ToolLoopAgent, MCP, embeddings, image gen, API routes), matching the 'fully executable code/commands' anchor.

3 / 3

Workflow Clarity

Sections and numbered Key Patterns give sequence, but the setup/agent/image flows lack explicit validation checkpoints for fragile or destructive operations (e.g., no verify step after image generation or gateway routing), capping it at the 'sequence present but checkpoints missing' level rather than 3.

2 / 3

Progressive Disclosure

The body is a large inline monolith; the one bundle file (references/providers.md) is not referenced or linked from the body, and substantial content (the full migration list, provider routing) that could live one level deep is kept inline. Cross-skill signals ("⤳ skill: ai-elements") help, but the single reference is not clearly signaled, matching the 'some structure but content that should be separate is inline' anchor.

2 / 3

Total

9

/

12

Passed

Description

100%

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 is strong: specific, third-person, and clearly triggers on natural user phrasing with an explicit 'Use when' clause. It answers both what the skill does and when to use it without vague fluff.

DimensionReasoningScore

Specificity

Lists many concrete capabilities — "chat interfaces, text generation, structured output, tool calling, agents, MCP integration, streaming, embeddings, reranking, image generation" — matching the 'lists multiple specific concrete actions' anchor.

3 / 3

Completeness

Explicitly answers both what ("Vercel AI SDK expert guidance" + enumerated features) and when ("Use when building AI-powered features ..."), satisfying the full what-and-when anchor.

3 / 3

Trigger Term Quality

Natural terms users would say ("chat interfaces", "tool calling", "streaming", "embeddings") plus an explicit "Use when building..." clause give good coverage of common phrasings.

3 / 3

Distinctiveness Conflict Risk

Scoped to the Vercel AI SDK with a distinct, enumerated trigger set, making it unlikely to fire for unrelated skills.

3 / 3

Total

12

/

12

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.

Validation13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (799 lines); consider splitting into references/ and linking

Warning

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

Total

13

/

16

Passed

Repository
openai/plugins
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.