chunking-embeddings

Chunking, embeddings, and RAG pipeline integration

Quality

68%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./.ai-rulez/skills/chunking-embeddings/SKILL.md

Quality

Content

76%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.

An exceptionally dense, high-signal reference for the codebase's chunking and embedding internals: it flags real traps (silent serde aliases, feature-gated presets, conflicting defaults) with no wasted tokens. The main gaps are the absence of an explicit ordered chunk→embed workflow with checkpoints and a copy-paste code example.

Suggestions

Add a short ordered workflow (configure ChunkingConfig with wire names → chunk_text/chunk_for_rag → embed per chunk) with a checkpoint to verify chunk_count and embedding presence before downstream use.

Include one minimal executable Rust snippet showing the common chunk-then-embed call so the entry points can be used verbatim.

Consider moving the preset table and wire-name table to a references/ file if the skill grows, keeping SKILL.md as the overview.

Dimension	Reasoning	Score
Conciseness	Every line carries non-obvious, load-bearing information — serde wire-name renames ("max_chars"/"max_overlap"), disagreeing defaults between EmbeddingModelType::default() and EmbeddingConfig::default(), the feature-gated no-op preset resolution, and the "no fastembed dependency" warning. There is no padding and no explanation of concepts Claude already knows.	5 / 5
Actionability	Provides concrete entry-point signatures (chunk_text, chunk_for_rag), a config field table with wire names and defaults, the full preset table with a source-of-truth pointer, and explicit do/don't Critical Rules. Falls short of a 5 only because there is no copy-paste-ready code example for the common chunk-then-embed case.	4 / 5
Workflow Clarity	The document is organized in execution order (chunking before embeddings, then rules) and the "Critical Rules" section works as a do/don't checklist, but the chunk → embed → integrate sequence is implicit rather than explicitly stepped, and there are no validation checkpoints. The skill is not a destructive or batch operation, so the cap-3 rule is not the binding constraint — the implicit sequence is.	3 / 5
Progressive Disclosure	Well-organized sections with one-level-deep pointers to three clearly signaled sibling skills (extraction-pipeline-patterns, config-loading-precedence, feature-flag-policy) and no nesting. The eight-row preset table and the wire-name table could arguably move to a reference file, which is the minor gap keeping this below 5; with no bundle files present, the single-page split is reasonable.	4 / 5
	Total	16 / 20 Passed

Description

61%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 names a clear, specific technical niche but is a bare capability list: it lacks action verbs and any "Use when..." trigger guidance. Adding an explicit trigger clause and a few natural synonyms would move it into the top anchors.

Suggestions

Add an explicit trigger clause, e.g. "Use when chunking text, generating embeddings, or wiring RAG pipelines against crates/xberg."

Rewrite as third-person verb phrases ("Splits text into chunks, generates ONNX/static embeddings, integrates RAG pipelines") to convey concrete actions rather than noun labels.

Include common synonyms users would say, such as "text splitting", "vector embeddings", and "semantic search", to broaden trigger coverage.

Dimension	Reasoning	Score
Specificity	Names three concrete technical capabilities ("Chunking, embeddings, and RAG pipeline integration") but as noun phrases with no action verbs, sitting between anchor 2's generic domain naming and anchor 4's list of specific actions. It is not vague fluff, but it describes domains rather than actions a user can act on.	3 / 5
Completeness	The "what" is clear (chunking, embeddings, RAG pipeline integration) but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3 per the judging guidelines.	3 / 5
Trigger Term Quality	"chunking", "embeddings", and "RAG" are natural terms a user would say when needing this skill, giving good keyword coverage. Missing common variations such as "text splitting", "vector embeddings", or "semantic search" that would round out coverage.	4 / 5
Distinctiveness Conflict Risk	The chunking/embeddings/RAG niche is distinct with domain-specific triggers, with only minor overlap risk against closely related skills such as text extraction or pipeline configuration skills.	4 / 5
	Total	14 / 20 Passed

Validation

93%

Warnings & errors only

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

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: xberg-io/xberg
Path: .ai-rulez/skills/chunking-embeddings/SKILL.md
Commit: fbf07e1

Reviewed: 12 hours ago

Table of Contents

Discovery Implementation Validation

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.