redis-search

Name: redis-search
Author: redis

334installs22forks

How to install redis-search?

redis/agent-skills/redis-search

$npx -y skills add redis/agent-skills --skill redis-search

Installs into the current project.

Summary

Redis Search guidance covering FT.CREATE schema design, field type selection (TEXT, TAG, NUMERIC, GEO, GEOSHAPE, VECTOR, JSON path), DIALECT 2 query syntax, FT.SEARCH / FT.AGGREGATE / FT.HYBRID command selection, vector similarity with HNSW or FLAT, hybrid retrieval combining lex

Files · 1

View on GitHub

SKILL.md

1	# Redis Search
2
3	Single source of guidance for Redis Search — the retrieval surface that spans lexical, numeric, geo, JSON-path, and vector queries. Vector fields are part of the same `FT.CREATE` machinery as TEXT/TAG/NUMERIC fields, and `FT.HYBRID` blends lexical and vector ranking in one command, so this skill covers them together.
4
5	## When to apply
6
7	- Creating, modifying, or reviewing a Redis Search index (`FT.CREATE`, `FT.ALTER`).
8	- Writing or optimizing `FT.SEARCH`, `FT.AGGREGATE`, or `FT.HYBRID` queries.
9	- Picking between `TEXT`, `TAG`, `NUMERIC`, `GEO`, `GEOSHAPE`, `VECTOR`, or JSON-path fields.
10	- Defining a `VECTOR` field, choosing HNSW vs FLAT, tuning HNSW parameters.
11	- Building a retrieval-augmented generation (RAG) pipeline.
12	- Rolling out a new index schema without downtime.
13	- Troubleshooting empty results, slow queries, or tokenization issues with `FT.EXPLAIN`, `FT.PROFILE`, `FT.INFO`.
14
15	## 1. Pick the right command
16
17	Three query commands. Reach for the narrowest one that fits.
18
19	\| Command \| When to use \| Mental model \| Minimum Redis \|
20	\|---\|---\|---\|---\|
21	\| FT.SEARCH \| Document retrieval, ranked or sorted. Best default. \| Returns matching docs directly. \| 2.0 (module) / 8.0 (built-in) \|
22	\| FT.AGGREGATE \| Faceting, computed fields, custom output shape, analytics. \| Declarative pipeline: `LOAD`, `APPLY`, `GROUPBY`, `REDUCE`, `SORTBY`. \| 2.0 / 8.0 \|
23	\| FT.HYBRID \| Blend lexical (BM25) with vector similarity, with configurable fusion. \| Pipeline with explicit `SEARCH` + `VSIM` legs and a `COMBINE` fusion stage. \| 8.4.0 \|
24
25	```
26	# FT.SEARCH — most common
27	FT.SEARCH idx:products "@category:{electronics} @price:[100 500]" LIMIT 0 20 RETURN 3 name price category
28
29	# FT.AGGREGATE — top categories by avg price
30	FT.AGGREGATE idx:products "*" GROUPBY 1 @category REDUCE AVG 1 @price AS avg_price SORTBY 2 @avg_price DESC
31
32	# FT.HYBRID (Redis ≥ 8.4) — lexical + vector fusion
33	FT.HYBRID idx:docs
34	SEARCH "@title:transformers" SCORER BM25 YIELD_SCORE_AS lexscore
35	VSIM embedding $vec KNN count 1 K 50 YIELD_SCORE_AS vecscore
36	COMBINE RRF 2 CONSTANT 60
37	PARAMS 2 vec "..."
38	DIALECT 2
39	```
40
41	For Redis < 8.4 the lexical+vector blend is approximated with `FT.SEARCH` pre-filter + `=>[KNN ...]`. See [references/command-selection.md](references/command-selection.md) and [references/hybrid-search.md](references/hybrid-search.md).
42
43	## 2. Schema basics — `FT.CREATE`
44
45	`FT.CREATE` indexes Hash or JSON documents matching a `PREFIX`. Always set `PREFIX`. Use `DIALECT 2` (the default since Redis 8; required for vector queries).
46
47	```
48	FT.CREATE idx:products ON HASH PREFIX 1 product:
49	SCHEMA
50	name TEXT WEIGHT 2.0
51	category TAG SORTABLE
52	price NUMERIC SORTABLE
53	location GEO
54	embedding VECTOR HNSW 6
55	TYPE FLOAT32
56	DIM 1536
57	DISTANCE_METRIC COSINE
58	```
59
60	Pick the narrowest field type that supports your access pattern:
61
62	\| Field type \| Use when \| Notes \|
63	\|---\|---\|---\|
64	\| `TEXT` \| Full-text search \| Tokenized + stemmed; not for exact match \|
65	\| `TAG` \| Exact match / filtering \| Add `SORTABLE UNF` for fastest tag queries \|
66	\| `NUMERIC` \| Range queries, sorting \| Prices, counts, timestamps \|
67	\| `GEO` \| Lat/long points \| Stores, users \|
68	\| `GEOSHAPE` \| Polygon / area queries \| Delivery zones, regions \|
69	\| `VECTOR` \| Similarity search \| HNSW or FLAT; see §4 \|
70	\| JSON `$.path AS alias` \| Nested JSON fields \| `ON JSON`; see [references/json-indexing.md](references/json-indexing.md) \|
71
72	The classic mistake is `TEXT` for a category or status field "because it's a string" — `TAG` is roughly 10× faster for exact-match filtering.
73
74	See [references/index-creation.md](references/index-creation.md), [references/field-types.md](references/field-types.md), [references/dialect.md](references/dialect.md), [references/ft-create-options.md](references/ft-create-options.md), [references/json-indexing.md](references/json-indexing.md).
75
76	## 3. Common queries
77
78	Narrow with filters; return only what you need.
79
80	```
81	# Tag filter + numeric range, sorted by price
82	FT.SEARCH idx:products "@category:{electronics} @price:[100 500]"
83	SORTBY price ASC
84	LIMIT 0 20
85	RETURN 3 name price category
86
87	# Text + tag filter
88	FT.SEARCH idx:products "wireless headphones @

People who liked this also viewed

6 picks

Type

redis-search

334installs22forks

View on GitHub

How to install redis-search?

redis/agent-skills/redis-search

$npx -y skills add redis/agent-skills --skill redis-search

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Redis Search
2
3	Single source of guidance for Redis Search — the retrieval surface that spans lexical, numeric, geo, JSON-path, and vector queries. Vector fields are part of the same `FT.CREATE` machinery as TEXT/TAG/NUMERIC fields, and `FT.HYBRID` blends lexical and vector ranking in one command, so this skill covers them together.
4
5	## When to apply
6
7	- Creating, modifying, or reviewing a Redis Search index (`FT.CREATE`, `FT.ALTER`).
8	- Writing or optimizing `FT.SEARCH`, `FT.AGGREGATE`, or `FT.HYBRID` queries.
9	- Picking between `TEXT`, `TAG`, `NUMERIC`, `GEO`, `GEOSHAPE`, `VECTOR`, or JSON-path fields.
10	- Defining a `VECTOR` field, choosing HNSW vs FLAT, tuning HNSW parameters.
11	- Building a retrieval-augmented generation (RAG) pipeline.
12	- Rolling out a new index schema without downtime.
13	- Troubleshooting empty results, slow queries, or tokenization issues with `FT.EXPLAIN`, `FT.PROFILE`, `FT.INFO`.
14
15	## 1. Pick the right command
16
17	Three query commands. Reach for the narrowest one that fits.
18
19	\| Command \| When to use \| Mental model \| Minimum Redis \|
20	\|---\|---\|---\|---\|
21	\| FT.SEARCH \| Document retrieval, ranked or sorted. Best default. \| Returns matching docs directly. \| 2.0 (module) / 8.0 (built-in) \|
22	\| FT.AGGREGATE \| Faceting, computed fields, custom output shape, analytics. \| Declarative pipeline: `LOAD`, `APPLY`, `GROUPBY`, `REDUCE`, `SORTBY`. \| 2.0 / 8.0 \|
23	\| FT.HYBRID \| Blend lexical (BM25) with vector similarity, with configurable fusion. \| Pipeline with explicit `SEARCH` + `VSIM` legs and a `COMBINE` fusion stage. \| 8.4.0 \|
24
25	```
26	# FT.SEARCH — most common
27	FT.SEARCH idx:products "@category:{electronics} @price:[100 500]" LIMIT 0 20 RETURN 3 name price category
28
29	# FT.AGGREGATE — top categories by avg price
30	FT.AGGREGATE idx:products "*" GROUPBY 1 @category REDUCE AVG 1 @price AS avg_price SORTBY 2 @avg_price DESC
31
32	# FT.HYBRID (Redis ≥ 8.4) — lexical + vector fusion
33	FT.HYBRID idx:docs
34	SEARCH "@title:transformers" SCORER BM25 YIELD_SCORE_AS lexscore
35	VSIM embedding $vec KNN count 1 K 50 YIELD_SCORE_AS vecscore
36	COMBINE RRF 2 CONSTANT 60
37	PARAMS 2 vec "..."
38	DIALECT 2
39	```
40
41	For Redis < 8.4 the lexical+vector blend is approximated with `FT.SEARCH` pre-filter + `=>[KNN ...]`. See [references/command-selection.md](references/command-selection.md) and [references/hybrid-search.md](references/hybrid-search.md).
42
43	## 2. Schema basics — `FT.CREATE`
44
45	`FT.CREATE` indexes Hash or JSON documents matching a `PREFIX`. Always set `PREFIX`. Use `DIALECT 2` (the default since Redis 8; required for vector queries).
46
47	```
48	FT.CREATE idx:products ON HASH PREFIX 1 product:
49	SCHEMA
50	name TEXT WEIGHT 2.0
51	category TAG SORTABLE
52	price NUMERIC SORTABLE
53	location GEO
54	embedding VECTOR HNSW 6
55	TYPE FLOAT32
56	DIM 1536
57	DISTANCE_METRIC COSINE
58	```
59
60	Pick the narrowest field type that supports your access pattern:
61
62	\| Field type \| Use when \| Notes \|
63	\|---\|---\|---\|
64	\| `TEXT` \| Full-text search \| Tokenized + stemmed; not for exact match \|
65	\| `TAG` \| Exact match / filtering \| Add `SORTABLE UNF` for fastest tag queries \|
66	\| `NUMERIC` \| Range queries, sorting \| Prices, counts, timestamps \|
67	\| `GEO` \| Lat/long points \| Stores, users \|
68	\| `GEOSHAPE` \| Polygon / area queries \| Delivery zones, regions \|
69	\| `VECTOR` \| Similarity search \| HNSW or FLAT; see §4 \|
70	\| JSON `$.path AS alias` \| Nested JSON fields \| `ON JSON`; see [references/json-indexing.md](references/json-indexing.md) \|
71
72	The classic mistake is `TEXT` for a category or status field "because it's a string" — `TAG` is roughly 10× faster for exact-match filtering.
73
74	See [references/index-creation.md](references/index-creation.md), [references/field-types.md](references/field-types.md), [references/dialect.md](references/dialect.md), [references/ft-create-options.md](references/ft-create-options.md), [references/json-indexing.md](references/json-indexing.md).
75
76	## 3. Common queries
77
78	Narrow with filters; return only what you need.
79
80	```
81	# Tag filter + numeric range, sorted by price
82	FT.SEARCH idx:products "@category:{electronics} @price:[100 500]"
83	SORTBY price ASC
84	LIMIT 0 20
85	RETURN 3 name price category
86
87	# Text + tag filter
88	FT.SEARCH idx:products "wireless headphones @

People who liked this also viewed

6 picks

Type