pinecone

Name: pinecone
Author: AlexAI-MCP

132stars24forks

How to install pinecone?

alexai-mcp/hermes-ccc/pinecone

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

Installs into the current project.

Summary

Managed vector database for production RAG — serverless and pod-based deployment, hybrid search, namespaces, and metadata filtering.

Files · 1

View on GitHub

SKILL.md

1	# Pinecone — Managed Vector Database
2
3	Fully managed vector database for production RAG. Serverless (pay-per-query) or pod-based (dedicated).
4
5	## Setup
6
7	```bash
8	pip install pinecone-client sentence-transformers openai
9	```
10
11	```python
12	from pinecone import Pinecone, ServerlessSpec
13
14	pc = Pinecone(api_key="your-api-key") # or os.environ["PINECONE_API_KEY"]
15	```
16
17	---
18
19	## Create Index
20
21	```python
22	# Serverless (pay-per-query — cheapest to start)
23	pc.create_index(
24	name="my-index",
25	dimension=1536, # match your embedding model
26	metric="cosine", # cosine \| euclidean \| dotproduct
27	spec=ServerlessSpec(
28	cloud="aws",
29	region="us-east-1"
30	)
31	)
32
33	# Connect to index
34	index = pc.Index("my-index")
35	```
36
37	---
38
39	## Upsert Vectors
40
41	```python
42	from sentence_transformers import SentenceTransformer
43
44	model = SentenceTransformer("all-MiniLM-L6-v2") # dim=384
45
46	documents = [
47	{"id": "doc1", "text": "Python async programming guide"},
48	{"id": "doc2", "text": "Machine learning with PyTorch"},
49	]
50
51	vectors = []
52	for doc in documents:
53	embedding = model.encode(doc["text"]).tolist()
54	vectors.append({
55	"id": doc["id"],
56	"values": embedding,
57	"metadata": {"text": doc["text"], "source": "manual"}
58	})
59
60	# Batch upsert (max 100 per call)
61	index.upsert(vectors=vectors, namespace="docs")
62	```
63
64	---
65
66	## Query
67
68	```python
69	query_text = "how to write async Python?"
70	query_vector = model.encode(query_text).tolist()
71
72	results = index.query(
73	vector=query_vector,
74	top_k=5,
75	namespace="docs",
76	include_metadata=True,
77	)
78
79	for match in results["matches"]:
80	print(f"Score: {match['score']:.3f} \| {match['metadata']['text']}")
81	```
82
83	---
84
85	## Metadata Filtering
86
87	```python
88	results = index.query(
89	vector=query_vector,
90	top_k=5,
91	filter={"source": {"$eq": "manual"}},
92	include_metadata=True,
93	)
94
95	# Operators: $eq, $ne, $gt, $gte, $lt, $lte, $in, $nin, $and, $or
96	results = index.query(
97	vector=query_vector,
98	top_k=5,
99	filter={
100	"$and": [
101	{"category": {"$in": ["tech", "science"]}},
102	{"year": {"$gte": 2023}},
103	]
104	},
105	include_metadata=True,
106	)
107	```
108
109	---
110
111	## Namespaces
112
113	```python
114	# Different namespaces = separate vector spaces (free, no extra cost)
115	index.upsert(vectors=vectors, namespace="user-123")
116	index.upsert(vectors=vectors, namespace="user-456")
117
118	# Query specific namespace
119	results = index.query(vector=query_vector, top_k=5, namespace="user-123")
120
121	# Delete namespace
122	index.delete(delete_all=True, namespace="user-123")
123	```
124
125	---
126
127	## Fetch / Delete / Update
128
129	```python
130	# Fetch specific vectors
131	fetched = index.fetch(ids=["doc1", "doc2"], namespace="docs")
132
133	# Delete vectors
134	index.delete(ids=["doc1"], namespace="docs")
135
136	# Update metadata (re-upsert with same id)
137	index.upsert(vectors=[{"id": "doc1", "values": embedding, "metadata": {"updated": True}}])
138	```
139
140	---
141
142	## Index Stats
143
144	```python
145	stats = index.describe_index_stats()
146	print(f"Total vectors: {stats['total_vector_count']}")
147	print(f"Namespaces: {stats['namespaces']}")
148	```
149
150	---
151
152	## When to Use Pinecone vs Alternatives
153
154	\| \| Pinecone \| Qdrant \| Chroma \|
155	\|---\|---\|---\|---\|
156	\| Hosting \| Managed cloud \| Self/cloud \| Self/cloud \|
157	\| Cost \| Pay-per-use \| Self-hosted free \| Free \|
158	\| Scale \| Billions \| Millions+ \| Millions \|
159	\| Setup \| Minutes \| Minutes \| Seconds \|
160	\| Best for \| Production SaaS \| Production self-hosted \| Local dev/RAG \|

People who liked this also viewed

6 picks

Type

find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skSkillsJul 20262.6M27k
grill-meA relentless interview to sharpen a plan or design.SkillsJul 2026633k182k
vercel-react-best-practicesVercel's official collection of agent skillsSkillsJul 2026573k29k
improve-codebase-architectureScan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.SkillsJul 2026522k182k
setup-matt-pocock-skillsConfigure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.SkillsJul 2026454k182k
triageMove issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.SkillsJul 2026424k182k

pinecone

132stars24forks

View on GitHub

How to install pinecone?

alexai-mcp/hermes-ccc/pinecone

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

Installs into the current project.

Summary

Managed vector database for production RAG — serverless and pod-based deployment, hybrid search, namespaces, and metadata filtering.

Files · 1

View on GitHub

SKILL.md

1	# Pinecone — Managed Vector Database
2
3	Fully managed vector database for production RAG. Serverless (pay-per-query) or pod-based (dedicated).
4
5	## Setup
6
7	```bash
8	pip install pinecone-client sentence-transformers openai
9	```
10
11	```python
12	from pinecone import Pinecone, ServerlessSpec
13
14	pc = Pinecone(api_key="your-api-key") # or os.environ["PINECONE_API_KEY"]
15	```
16
17	---
18
19	## Create Index
20
21	```python
22	# Serverless (pay-per-query — cheapest to start)
23	pc.create_index(
24	name="my-index",
25	dimension=1536, # match your embedding model
26	metric="cosine", # cosine \| euclidean \| dotproduct
27	spec=ServerlessSpec(
28	cloud="aws",
29	region="us-east-1"
30	)
31	)
32
33	# Connect to index
34	index = pc.Index("my-index")
35	```
36
37	---
38
39	## Upsert Vectors
40
41	```python
42	from sentence_transformers import SentenceTransformer
43
44	model = SentenceTransformer("all-MiniLM-L6-v2") # dim=384
45
46	documents = [
47	{"id": "doc1", "text": "Python async programming guide"},
48	{"id": "doc2", "text": "Machine learning with PyTorch"},
49	]
50
51	vectors = []
52	for doc in documents:
53	embedding = model.encode(doc["text"]).tolist()
54	vectors.append({
55	"id": doc["id"],
56	"values": embedding,
57	"metadata": {"text": doc["text"], "source": "manual"}
58	})
59
60	# Batch upsert (max 100 per call)
61	index.upsert(vectors=vectors, namespace="docs")
62	```
63
64	---
65
66	## Query
67
68	```python
69	query_text = "how to write async Python?"
70	query_vector = model.encode(query_text).tolist()
71
72	results = index.query(
73	vector=query_vector,
74	top_k=5,
75	namespace="docs",
76	include_metadata=True,
77	)
78
79	for match in results["matches"]:
80	print(f"Score: {match['score']:.3f} \| {match['metadata']['text']}")
81	```
82
83	---
84
85	## Metadata Filtering
86
87	```python
88	results = index.query(
89	vector=query_vector,
90	top_k=5,
91	filter={"source": {"$eq": "manual"}},
92	include_metadata=True,
93	)
94
95	# Operators: $eq, $ne, $gt, $gte, $lt, $lte, $in, $nin, $and, $or
96	results = index.query(
97	vector=query_vector,
98	top_k=5,
99	filter={
100	"$and": [
101	{"category": {"$in": ["tech", "science"]}},
102	{"year": {"$gte": 2023}},
103	]
104	},
105	include_metadata=True,
106	)
107	```
108
109	---
110
111	## Namespaces
112
113	```python
114	# Different namespaces = separate vector spaces (free, no extra cost)
115	index.upsert(vectors=vectors, namespace="user-123")
116	index.upsert(vectors=vectors, namespace="user-456")
117
118	# Query specific namespace
119	results = index.query(vector=query_vector, top_k=5, namespace="user-123")
120
121	# Delete namespace
122	index.delete(delete_all=True, namespace="user-123")
123	```
124
125	---
126
127	## Fetch / Delete / Update
128
129	```python
130	# Fetch specific vectors
131	fetched = index.fetch(ids=["doc1", "doc2"], namespace="docs")
132
133	# Delete vectors
134	index.delete(ids=["doc1"], namespace="docs")
135
136	# Update metadata (re-upsert with same id)
137	index.upsert(vectors=[{"id": "doc1", "values": embedding, "metadata": {"updated": True}}])
138	```
139
140	---
141
142	## Index Stats
143
144	```python
145	stats = index.describe_index_stats()
146	print(f"Total vectors: {stats['total_vector_count']}")
147	print(f"Namespaces: {stats['namespaces']}")
148	```
149
150	---
151
152	## When to Use Pinecone vs Alternatives
153
154	\| \| Pinecone \| Qdrant \| Chroma \|
155	\|---\|---\|---\|---\|
156	\| Hosting \| Managed cloud \| Self/cloud \| Self/cloud \|
157	\| Cost \| Pay-per-use \| Self-hosted free \| Free \|
158	\| Scale \| Billions \| Millions+ \| Millions \|
159	\| Setup \| Minutes \| Minutes \| Seconds \|
160	\| Best for \| Production SaaS \| Production self-hosted \| Local dev/RAG \|

People who liked this also viewed

6 picks

Type