CtrlK
BlogDocsLog inGet started
Tessl Logo

clickhouse-io

ClickHouse database patterns, query optimization, analytics, and data engineering best practices for high-performance analytical workloads.

57

Quality

65%

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 ./docs/zh-TW/skills/clickhouse-io/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

65%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 dense, highly actionable pattern catalog with strong copy-paste SQL/TypeScript examples and effective PASS/FAIL contrasts. Its weaknesses are structural: no bundle-file split for a 430-line reference, a generic overview section explaining ClickHouse basics Claude already knows, and no validation/verification checkpoints for the batch insert and ETL/CDC workflows it teaches.

Suggestions

Split the ~430-line catalog into one-level-deep reference files (e.g., ANALYTICS_QUERIES.md, DATA_PIPELINES.md, MONITORING.md), keeping SKILL.md as a concise overview with clearly signaled links.

Add validation/feedback checkpoints to the batch and pipeline workflows (e.g., verify insert counts after bulk load, check system.query_log for failures before/after ETL runs) — their absence currently caps workflow clarity at 3.

Trim the '概述' section explaining what ClickHouse/OLAP and columnar storage are (concepts Claude already knows) and remove the closing '記住' summary line; also fix the CDC example to batch notifications instead of single-row inserts, which contradicts the skill's own batch-insert best practice.

DimensionReasoningScore

Conciseness

The body is dominated by lean, copy-paste-ready SQL/TypeSQL examples with minimal prose, which assumes Claude's competence. However, the '概述' section ("ClickHouse 是一個列式資料庫管理系統(DBMS),用於線上分析處理(OLAP)" plus the '關鍵特性' list of columnar storage/compression/parallel execution) explains concepts Claude already knows, and the closing "記住" line adds no new information. This is minor over-explanation trimmable to a tighter form, fitting score 4 rather than 5, and far above the padded verbosity of scores 1–2.

4 / 5

Actionability

Nearly all guidance is concrete executable code — MergeTree/AggregatingMergeTree DDL, PASS/FAIL query contrasts, bulk insert code, materialized view DDL, system.query_log monitoring queries, retention/funnel/cohort recipes. It falls short of 5 because several examples depend on undefined scaffolding (bulkInsertToClickHouse, extractFromPostgres, the Trade type, the events table's columns), and the CDC example inserts one row per notification, directly contradicting the skill's own batch-insert best practice.

4 / 5

Workflow Clarity

The content is a well-organized pattern catalog rather than a sequenced workflow, and the batch/data-pipeline operations it teaches (bulk inserts, ETL, CDC sync) have no validation or verification checkpoints — the monitoring section shows how to inspect slow queries but no validate → fix → retry loop exists. Per the rubric, batch operations without validation cap workflow clarity at 3; it is not 2 because material is logically organized and the PASS/FAIL examples do communicate correct vs. incorrect approaches.

3 / 5

Progressive Disclosure

The ~430-line body has clear section headers and no broken or nested references (there are no bundle files at all), but everything — analytics query recipes, ETL/CDC pipeline patterns, monitoring SQL — is inlined in SKILL.md where a reference catalog of this size would better be split into one-level-deep reference files. This matches score 3: good internal structure but content that should be separate is inline; not 4–5 because there is no overview-plus-references split at all.

3 / 5

Total

14

/

20

Passed

Description

65%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 clearly identifies the ClickHouse niche with good natural trigger terms, but it reads as a topic list rather than a list of concrete actions, and it entirely lacks a 'Use when...' trigger clause, capping completeness. Adding explicit activation conditions and replacing buzzwords ('best practices', 'high-performance analytical workloads') with concrete actions would lift it substantially.

Suggestions

Add an explicit 'Use when...' trigger clause (e.g., 'Use when writing or optimizing ClickHouse SQL, designing MergeTree tables, or tuning slow analytical queries'), which currently caps the completeness dimension at 3.

Replace buzzword phrases like 'data engineering best practices' and 'high-performance analytical workloads' with concrete actions (e.g., 'designs MergeTree tables, builds materialized views for real-time aggregation, tunes slow queries via system.query_log').

Add natural trigger variations users would say, such as 'OLAP', 'SQL', 'slow query', or 'performance tuning', to strengthen keyword coverage toward score 5.

DimensionReasoningScore

Specificity

The description names the domain ("ClickHouse database patterns, query optimization") and a couple of concrete capability areas, but "analytics", "data engineering best practices", and "high-performance analytical workloads" are topic labels and buzzwords rather than concrete actions, so coverage is not comprehensive. It sits above score 2 (which names the domain but with minimal/generic actions) and below score 4 (which would list several specific actions like 'extract text, fill forms, merge documents') because no concrete verbs or deliverables are enumerated.

3 / 5

Completeness

The 'what' is reasonably clear (ClickHouse patterns, query optimization, analytics, data engineering best practices), but there is no 'Use when...' or equivalent trigger clause telling Claude when to load the skill. Per the judging guidelines, a missing explicit trigger clause caps completeness at 3; it is not 4 or 5 because no 'when' is present even weakly.

3 / 5

Trigger Term Quality

Natural terms a user would actually say are present — "ClickHouse", "query optimization", "analytics", "data engineering" — giving good keyword coverage. A few natural variations are missing (e.g., "SQL", "OLAP", "slow query", "performance tuning", "materialized views"), which keeps it below score 5's comprehensive synonym/extension coverage, but it is clearly above score 3's 'some relevant keywords missing common variations' since the primary product name plus the main task phrases all appear.

4 / 5

Distinctiveness Conflict Risk

"ClickHouse" is a specific named technology, giving the skill a clear niche with distinct triggers and minimal conflict risk (only slight overlap with generic SQL/database skills). It matches the score 5 anchor of a clear niche with distinct triggers.

5 / 5

Total

15

/

20

Passed

Validation

100%

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
affaan-m/ECC
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.