chroma

Name: chroma
Author: AlexAI-MCP

132stars24forks

How to install chroma?

alexai-mcp/hermes-ccc/chroma

$npx -y skills add AlexAI-MCP/hermes-CCC --skill chroma

Installs into the current project.

Summary

Open-source embedding database for RAG — store embeddings, vector search, metadata filtering. Simple API, scales from notebook to production.

Files · 1

View on GitHub

SKILL.md

1	# Chroma
2
3	## Purpose
4
5	- Use this skill to store embeddings and perform semantic retrieval for RAG systems.
6	- Prefer Chroma when you want a simple local developer experience with an easy Python API.
7	- Chroma works well for notebooks, prototypes, local apps, and moderate production deployments.
8	- It supports metadata filtering, server mode, and pluggable embedding functions.
9
10	## Install
11
12	```bash
13	pip install chromadb sentence-transformers
14	```
15
16	- Add any extra embedding provider packages you need, such as the OpenAI SDK.
17
18	## Basic Setup
19
20	```python
21	import chromadb
22
23	client = chromadb.PersistentClient(path="./chroma_db")
24	```
25
26	- `PersistentClient` stores data on disk.
27	- It is a good default for local development and single-node setups.
28
29	## Create a Collection
30
31	```python
32	collection = client.create_collection(name="docs")
33	```
34
35	- Use descriptive collection names like `docs`, `papers`, `tickets`, or `kb_chunks`.
36	- A collection is the logical container for your embeddings and payload metadata.
37
38	## Add Documents
39
40	- Minimal add call:
41
42	```python
43	collection.add(
44	documents=["Chroma is useful for local RAG.", "Vector search retrieves semantically similar text."],
45	metadatas=[{"source": "note1"}, {"source": "note2"}],
46	ids=["doc-1", "doc-2"],
47	)
48	```
49
50	- Required arrays must align by index.
51	- Keep `ids` stable if you plan to update or delete records later.
52
53	## Query
54
55	```python
56	results = collection.query(
57	query_texts=["How do I store embeddings for RAG?"],
58	n_results=5,
59	)
60
61	print(results["documents"])
62	print(results["metadatas"])
63	```
64
65	- `query_texts=['...']` is the most common path when Chroma is managing embeddings for you.
66	- Start with `n_results=5` or `10` for most retrieval experiments.
67
68	## Basic Collection Lifecycle
69
70	- Create collection
71	- Add documents
72	- Query for nearest neighbors
73	- Update documents when content changes
74	- Delete stale records
75
76	## Get or Create
77
78	- For idempotent startup flows, prefer `get_or_create_collection`:
79
80	```python
81	collection = client.get_or_create_collection(name="docs")
82	```
83
84	- This is helpful in services that initialize storage on boot.
85
86	## Embedding Functions
87
88	- Chroma supports multiple embedding strategies.
89	- Common choices include:
90	- `DefaultEmbeddingFunction`
91	- `SentenceTransformerEmbeddingFunction`
92	- `OpenAIEmbeddingFunction`
93
94	## Default Embedding Function
95
96	```python
97	from chromadb.utils.embedding_functions import DefaultEmbeddingFunction
98
99	embedding_fn = DefaultEmbeddingFunction()
100	collection = client.get_or_create_collection(
101	name="default-embeddings",
102	embedding_function=embedding_fn,
103	)
104	```
105
106	- The default function is convenient for quick experiments.
107	- For production, you usually want to control the embedding model explicitly.
108
109	## SentenceTransformers Embedding Function
110
111	```python
112	from chromadb.utils.embedding_functions import SentenceTransformerEmbeddingFunction
113
114	embedding_fn = SentenceTransformerEmbeddingFunction(
115	model_name="sentence-transformers/all-MiniLM-L6-v2",
116	)
117
118	collection = client.get_or_create_collection(
119	name="st-docs",
120	embedding_function=embedding_fn,
121	)
122	```
123
124	- This is a good choice for local semantic search with no external API dependency.
125
126	## OpenAI Embedding Function
127
128	```python
129	from chromadb.utils.embedding_functions import OpenAIEmbeddingFunction
130
131	embedding_fn = OpenAIEmbeddingFunction(
132	api_key="YOUR_API_KEY",
133	model_name="text-embedding-3-small",
134	)
135
136	collection = client.get_or_create_collection(
137	name="openai-docs",
138	embedding_function=embedding_fn,
139	)
140	```
141
142	- Use this when you want consistent hosted embeddings across environments.
143	- Track model versions because embedding changes can invalidate similarity behavior.
144
145	## Add Structured Documents
146
147	```python
148	collection.add(
149	documents=[
150	"Retrieval augmented generation combines retrieval with generation.",
151	"Chroma collections can store metadata for filtering.",
152	],
153	metadatas=[
154	{"source": "paper1", "section": "intro", "year": 2024},
155	{"source": "paper1", "section": "methods", "year": 2024},
156	],
157	ids=["paper1-intro", "paper1-methods"],
158	)
159	```
160
161	- Include source, section, title, author, or timestamp metadata if you need filtered retrieval.
162
163	## Metadata Filtering
164
165	- Filter by metadata with `where`:
166
167	```python
168	results = collection.query(
169	query_texts=["What does the paper say about filtering?"],
170	n_results=5,
171	where={"source": "paper1"},
172	)
173	```
174
175	- Example requested pattern:
176	- `where={'source': 'paper1'}`
177
178	- Metadata filters are critical for multi-tenant or source-restricted RAG.
179
180	## Full Text Search
181
182	- Chroma also supports document-side text filtering with `

People who liked this also viewed

6 picks

Type

azure-aiUse for Azure AI: Search, Speech, OpenAI, Document Intelligence. Helps with search, vector/hybrid search, speech-to-text, text-to-speech, transcription, OCR. WHEN: AI Search, query search, vector searSkillsJul 2026472k1.3k
just-scrapeSearch, scrape, crawl, extract structured data, and monitor web pages via the ScrapeGraph AI CLI. Use when the user asks to search the web, scrape a webpage, grab content from a URL, extract JSON fromSkillsJul 2026245k37
firecrawlCLI and Agent Skill for Firecrawl - Add scrape, search, and browsing capabilities to your AI agentsSkillsJul 202694k531
firecrawl-searchWeb search with full page content extraction. Use this skill whenever the user asks to search the web, find articles, research a topic, look something up, find recent news, discover sources, or says "SkillsJul 202676k531
firecrawl-agentAI-powered autonomous data extraction that navigates complex sites and returns structured JSON. Use this skill when the user wants structured data from websites, needs to extract pricing tiers, producSkillsJul 202674k531
firecrawl-mapDiscover and list all URLs on a website, with optional search filtering. Use this skill when the user wants to find a specific page on a large site, list all URLs, see the site structure, find where sSkillsJul 202674k531

chroma

132stars24forks

View on GitHub

How to install chroma?

alexai-mcp/hermes-ccc/chroma

$npx -y skills add AlexAI-MCP/hermes-CCC --skill chroma

Installs into the current project.

Summary

Open-source embedding database for RAG — store embeddings, vector search, metadata filtering. Simple API, scales from notebook to production.

Files · 1

View on GitHub

SKILL.md

1	# Chroma
2
3	## Purpose
4
5	- Use this skill to store embeddings and perform semantic retrieval for RAG systems.
6	- Prefer Chroma when you want a simple local developer experience with an easy Python API.
7	- Chroma works well for notebooks, prototypes, local apps, and moderate production deployments.
8	- It supports metadata filtering, server mode, and pluggable embedding functions.
9
10	## Install
11
12	```bash
13	pip install chromadb sentence-transformers
14	```
15
16	- Add any extra embedding provider packages you need, such as the OpenAI SDK.
17
18	## Basic Setup
19
20	```python
21	import chromadb
22
23	client = chromadb.PersistentClient(path="./chroma_db")
24	```
25
26	- `PersistentClient` stores data on disk.
27	- It is a good default for local development and single-node setups.
28
29	## Create a Collection
30
31	```python
32	collection = client.create_collection(name="docs")
33	```
34
35	- Use descriptive collection names like `docs`, `papers`, `tickets`, or `kb_chunks`.
36	- A collection is the logical container for your embeddings and payload metadata.
37
38	## Add Documents
39
40	- Minimal add call:
41
42	```python
43	collection.add(
44	documents=["Chroma is useful for local RAG.", "Vector search retrieves semantically similar text."],
45	metadatas=[{"source": "note1"}, {"source": "note2"}],
46	ids=["doc-1", "doc-2"],
47	)
48	```
49
50	- Required arrays must align by index.
51	- Keep `ids` stable if you plan to update or delete records later.
52
53	## Query
54
55	```python
56	results = collection.query(
57	query_texts=["How do I store embeddings for RAG?"],
58	n_results=5,
59	)
60
61	print(results["documents"])
62	print(results["metadatas"])
63	```
64
65	- `query_texts=['...']` is the most common path when Chroma is managing embeddings for you.
66	- Start with `n_results=5` or `10` for most retrieval experiments.
67
68	## Basic Collection Lifecycle
69
70	- Create collection
71	- Add documents
72	- Query for nearest neighbors
73	- Update documents when content changes
74	- Delete stale records
75
76	## Get or Create
77
78	- For idempotent startup flows, prefer `get_or_create_collection`:
79
80	```python
81	collection = client.get_or_create_collection(name="docs")
82	```
83
84	- This is helpful in services that initialize storage on boot.
85
86	## Embedding Functions
87
88	- Chroma supports multiple embedding strategies.
89	- Common choices include:
90	- `DefaultEmbeddingFunction`
91	- `SentenceTransformerEmbeddingFunction`
92	- `OpenAIEmbeddingFunction`
93
94	## Default Embedding Function
95
96	```python
97	from chromadb.utils.embedding_functions import DefaultEmbeddingFunction
98
99	embedding_fn = DefaultEmbeddingFunction()
100	collection = client.get_or_create_collection(
101	name="default-embeddings",
102	embedding_function=embedding_fn,
103	)
104	```
105
106	- The default function is convenient for quick experiments.
107	- For production, you usually want to control the embedding model explicitly.
108
109	## SentenceTransformers Embedding Function
110
111	```python
112	from chromadb.utils.embedding_functions import SentenceTransformerEmbeddingFunction
113
114	embedding_fn = SentenceTransformerEmbeddingFunction(
115	model_name="sentence-transformers/all-MiniLM-L6-v2",
116	)
117
118	collection = client.get_or_create_collection(
119	name="st-docs",
120	embedding_function=embedding_fn,
121	)
122	```
123
124	- This is a good choice for local semantic search with no external API dependency.
125
126	## OpenAI Embedding Function
127
128	```python
129	from chromadb.utils.embedding_functions import OpenAIEmbeddingFunction
130
131	embedding_fn = OpenAIEmbeddingFunction(
132	api_key="YOUR_API_KEY",
133	model_name="text-embedding-3-small",
134	)
135
136	collection = client.get_or_create_collection(
137	name="openai-docs",
138	embedding_function=embedding_fn,
139	)
140	```
141
142	- Use this when you want consistent hosted embeddings across environments.
143	- Track model versions because embedding changes can invalidate similarity behavior.
144
145	## Add Structured Documents
146
147	```python
148	collection.add(
149	documents=[
150	"Retrieval augmented generation combines retrieval with generation.",
151	"Chroma collections can store metadata for filtering.",
152	],
153	metadatas=[
154	{"source": "paper1", "section": "intro", "year": 2024},
155	{"source": "paper1", "section": "methods", "year": 2024},
156	],
157	ids=["paper1-intro", "paper1-methods"],
158	)
159	```
160
161	- Include source, section, title, author, or timestamp metadata if you need filtered retrieval.
162
163	## Metadata Filtering
164
165	- Filter by metadata with `where`:
166
167	```python
168	results = collection.query(
169	query_texts=["What does the paper say about filtering?"],
170	n_results=5,
171	where={"source": "paper1"},
172	)
173	```
174
175	- Example requested pattern:
176	- `where={'source': 'paper1'}`
177
178	- Metadata filters are critical for multi-tenant or source-restricted RAG.
179
180	## Full Text Search
181
182	- Chroma also supports document-side text filtering with `

People who liked this also viewed

6 picks

Type