CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

chunking

Use when splitting extracted text into chunks for LLM context windows or RAG ingestion. Covers chunk size, overlap, markdown/yaml/semantic chunkers, tokenizer-based sizing, and the standalone `chunk` command.

67

Quality

84%

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

The canonical home for this skill is chunking in xberg-io/xberg

SKILL.md
Quality
Evals
Security

Quality

Content

78%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 highly actionable, well-organized body with executable examples for every usage mode and a useful parameter-selection guide. The main defect is progressive disclosure: the two referenced files do not exist in the bundle, and API-naming minutiae are inlined instead of living in those references.

Suggestions

Create the referenced files (references/configuration.md and references/cli-reference.md) or remove the pointers — as shipped, both references are dangling paths.

Move the cross-language field-naming note (Python/Rust/TOML/Node aliases) into the API reference files and keep only the snake_case config example inline.

Trim the chunker_type capitalization explanation to the actionable rule (lowercase before comparing) and drop the Rust Debug backstory.

DimensionReasoningScore

Conciseness

Largely lean — terse commands, a comparison table, compact pitfalls — with minor over-explanation that could be trimmed, such as the Rust Debug capitalization aside and the dense cross-language field-naming paragraph.

4 / 5

Actionability

Fully executable, copy-paste-ready bash, TOML, and Python examples cover the common cases: inline extraction chunking, standalone chunk, markdown/semantic chunkers, tokenizer sizing, config-file, and programmatic access.

5 / 5

Workflow Clarity

Content is organized by usage mode with a 'Picking parameters' decision guide and a pitfalls section covering failure modes (overlap >= size rejection, missing feature, empty input), but there is no explicit verify step such as checking the resolved config in output.

4 / 5

Progressive Disclosure

References are clearly signaled and one level deep ('See `references/configuration.md` ... and `references/cli-reference.md`'), but no references/ directory exists in this bundle — both pointers are dangling — and the cross-language naming-alias note is inline detail that belongs in those reference files.

3 / 5

Total

16

/

20

Passed

Description

83%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 strong description with an explicit 'Use when' trigger, concrete capability coverage, and good natural keywords. Its only weaknesses are a capability-list framing rather than concrete actions and a few missing synonyms (vector store, embeddings, split text).

DimensionReasoningScore

Specificity

Names the domain ('splitting extracted text into chunks') plus several concrete facets — 'chunk size, overlap, markdown/yaml/semantic chunkers, tokenizer-based sizing, and the standalone `chunk` command' — but the coverage list is parameter-shaped rather than action-shaped, leaving minor gaps.

4 / 5

Completeness

Explicitly answers both: what ('Covers chunk size, overlap, markdown/yaml/semantic chunkers, tokenizer-based sizing, and the standalone `chunk` command') and when ('Use when splitting extracted text into chunks for LLM context windows or RAG ingestion') with a concrete trigger phrase.

5 / 5

Trigger Term Quality

Includes natural terms users would say ('chunks', 'RAG', 'LLM context windows', 'chunk size', 'overlap') but misses common variations like 'vector store', 'embeddings', 'split text', and 'text segmentation'.

4 / 5

Distinctiveness Conflict Risk

'Chunking', 'RAG ingestion', and the '`chunk` command' carve a clear niche, but 'LLM context windows' and 'RAG ingestion' brush against adjacent ingestion/vector-store skills, creating minor overlap risk.

4 / 5

Total

17

/

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

referenced_paths_exist

Referenced path issues: 4 missing

Warning

Total

15

/

16

Passed

Repository

xberg-io/xberg

Path

plugin/.hermes/package/src/xberg_hermes_plugin/skills/chunking/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.