clickhouse-architecture-advisor

Name: clickhouse-architecture-advisor
Author: ClickHouse

4.1kinstalls32forks

How to install clickhouse-architecture-advisor?

clickhouse/agent-skills/clickhouse-architecture-advisor

$npx -y skills add clickhouse/agent-skills --skill clickhouse-architecture-advisor

Installs into the current project.

Summary

MUST USE when designing ClickHouse architectures, selecting between ingestion or modeling patterns, or translating best practices into workload-specific system designs. Complements clickhouse-best-practices with decision frameworks and explicit provenance labels.

Files · 1

View on GitHub

SKILL.md

1	# ClickHouse Architecture Advisor
2
3	This skill adds workload-aware architecture decisioning on top of `clickhouse-best-practices`.
4
5	> Official docs remain the source of truth.
6	> This skill must always prefer official ClickHouse documentation when available.
7
8	## Required behavior
9
10	Before producing recommendations:
11
12	1. Identify the workload shape
13	- observability
14	- security / SIEM
15	- product analytics
16	- IoT / telemetry
17	- market data / financial services
18	- mixed OLAP with point-lookups
19	2. Read the relevant decision rule files in `rules/`
20	3. Use `mappings/doc_links.yaml` to attach official documentation
21	4. Classify every recommendation as:
22	- `official`
23	- `derived`
24	- `field`
25	5. Never present field guidance as official guidance
26	6. If a recommendation is uncertain, say so explicitly
27
28	## Provenance rules
29
30	### `official`
31	Use this when the recommendation is directly backed by official docs.
32
33	### `derived`
34	Use this when the recommendation is not stated verbatim in docs but follows logically from documented ClickHouse behavior.
35
36	### `field`
37	Use this only for experience-based guidance that may be situational.
38	When using `field`, include:
39	- a disclaimer that the advice is heuristic
40	- a relevant official doc if one partially applies
41	- the reason the advice depends on workload context
42
43	## Read these rule files by scenario
44
45	### Real-time ingestion design
46	1. `rules/decision-ingestion-strategy.md`
47	2. `rules/decision-real-time-preaggregation.md`
48	3. Relevant best-practices insert rules
49
50	### Time-series and retention design
51	1. `rules/decision-partitioning-timeseries.md`
52	2. Relevant best-practices schema partition rules
53
54	### Enrichment and dimension lookups
55	1. `rules/decision-join-enrichment.md`
56	2. Relevant best-practices query join rules
57
58	### Mutable state / late-arriving events
59	1. `rules/decision-late-arriving-upserts.md`
60	2. Relevant best-practices mutation avoidance rules
61
62	## Output format
63
64	Structure responses like this:
65
66	```markdown
67	## Workload Summary
68	- workload:
69	- latency target:
70	- data shape:
71	- primary query patterns:
72	- operational constraints:
73
74	## Key Decisions
75	- ...
76	- ...
77
78	## Recommendations
79
80	### <Recommendation title>
81
82	What
83	...
84
85	Why
86	...
87
88	How
89	...
90
91	Category
92	official \| derived \| field
93
94	Confidence
95	high \| medium \| heuristic
96
97	Source
98	- doc link(s)
99
100	Validation
101	- concrete SQL, metric, or smoke test
102	```
103
104	## Architecture-specific guidance
105
106	Prefer decision frameworks over generic advice. Good responses should:
107	- explain tradeoffs
108	- identify the likely operating bottleneck
109	- separate immediate actions from structural redesign
110	- provide target architecture patterns, not just isolated settings
111
112	## Full reference
113
114	See `AGENTS.md` for the compiled version and `examples/` for sample outputs.

People who liked this also viewed

6 picks

Type

clickhouse-architecture-advisor

4.1kinstalls32forks

View on GitHub

How to install clickhouse-architecture-advisor?

clickhouse/agent-skills/clickhouse-architecture-advisor

$npx -y skills add clickhouse/agent-skills --skill clickhouse-architecture-advisor

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# ClickHouse Architecture Advisor
2
3	This skill adds workload-aware architecture decisioning on top of `clickhouse-best-practices`.
4
5	> Official docs remain the source of truth.
6	> This skill must always prefer official ClickHouse documentation when available.
7
8	## Required behavior
9
10	Before producing recommendations:
11
12	1. Identify the workload shape
13	- observability
14	- security / SIEM
15	- product analytics
16	- IoT / telemetry
17	- market data / financial services
18	- mixed OLAP with point-lookups
19	2. Read the relevant decision rule files in `rules/`
20	3. Use `mappings/doc_links.yaml` to attach official documentation
21	4. Classify every recommendation as:
22	- `official`
23	- `derived`
24	- `field`
25	5. Never present field guidance as official guidance
26	6. If a recommendation is uncertain, say so explicitly
27
28	## Provenance rules
29
30	### `official`
31	Use this when the recommendation is directly backed by official docs.
32
33	### `derived`
34	Use this when the recommendation is not stated verbatim in docs but follows logically from documented ClickHouse behavior.
35
36	### `field`
37	Use this only for experience-based guidance that may be situational.
38	When using `field`, include:
39	- a disclaimer that the advice is heuristic
40	- a relevant official doc if one partially applies
41	- the reason the advice depends on workload context
42
43	## Read these rule files by scenario
44
45	### Real-time ingestion design
46	1. `rules/decision-ingestion-strategy.md`
47	2. `rules/decision-real-time-preaggregation.md`
48	3. Relevant best-practices insert rules
49
50	### Time-series and retention design
51	1. `rules/decision-partitioning-timeseries.md`
52	2. Relevant best-practices schema partition rules
53
54	### Enrichment and dimension lookups
55	1. `rules/decision-join-enrichment.md`
56	2. Relevant best-practices query join rules
57
58	### Mutable state / late-arriving events
59	1. `rules/decision-late-arriving-upserts.md`
60	2. Relevant best-practices mutation avoidance rules
61
62	## Output format
63
64	Structure responses like this:
65
66	```markdown
67	## Workload Summary
68	- workload:
69	- latency target:
70	- data shape:
71	- primary query patterns:
72	- operational constraints:
73
74	## Key Decisions
75	- ...
76	- ...
77
78	## Recommendations
79
80	### <Recommendation title>
81
82	What
83	...
84
85	Why
86	...
87
88	How
89	...
90
91	Category
92	official \| derived \| field
93
94	Confidence
95	high \| medium \| heuristic
96
97	Source
98	- doc link(s)
99
100	Validation
101	- concrete SQL, metric, or smoke test
102	```
103
104	## Architecture-specific guidance
105
106	Prefer decision frameworks over generic advice. Good responses should:
107	- explain tradeoffs
108	- identify the likely operating bottleneck
109	- separate immediate actions from structural redesign
110	- provide target architecture patterns, not just isolated settings
111
112	## Full reference
113
114	See `AGENTS.md` for the compiled version and `examples/` for sample outputs.

People who liked this also viewed

6 picks

Type