claude-api

Name: claude-api
Author: tranhieutt

69stars41forks

How to install claude-api?

tranhieutt/software_development_department/claude-api

$npx -y skills add tranhieutt/software_development_department --skill claude-api

Installs into the current project.

Summary

Provides code patterns for the Anthropic Claude API including streaming, tool use, and prompt caching. Use when working with Anthropic SDK files or when the user mentions Claude API, Anthropic client, or LLM integration.

Files · 1

View on GitHub

SKILL.md

1	# Claude API
2
3	Build applications with the Anthropic Claude API and SDKs.
4
5	## When to Activate
6
7	- Building applications that call the Claude API
8	- Code imports `anthropic` (Python) or `@anthropic-ai/sdk` (TypeScript)
9	- User asks about Claude API patterns, tool use, streaming, or vision
10	- Implementing agent workflows with Claude Agent SDK
11	- Optimizing API costs, token usage, or latency
12
13	## Model Selection
14
15	\| Model \| ID \| Best For \|
16	\|-------\|-----\|----------\|
17	\| Opus 4.1 \| `claude-opus-4-1` \| Complex reasoning, architecture, research \|
18	\| Sonnet 4 \| `claude-sonnet-4-0` \| Balanced coding, most development tasks \|
19	\| Haiku 3.5 \| `claude-3-5-haiku-latest` \| Fast responses, high-volume, cost-sensitive \|
20
21	Default to Sonnet 4 unless the task requires deep reasoning (Opus) or speed/cost optimization (Haiku). For production, prefer pinned snapshot IDs over aliases.
22
23	## Python SDK
24
25	### Installation
26
27	```bash
28	pip install anthropic
29	```
30
31	### Basic Message
32
33	```python
34	import anthropic
35
36	client = anthropic.Anthropic() # reads ANTHROPIC_API_KEY from env
37
38	message = client.messages.create(
39	model="claude-sonnet-4-0",
40	max_tokens=1024,
41	messages=[
42	{"role": "user", "content": "Explain async/await in Python"}
43	]
44	)
45	print(message.content[0].text)
46	```
47
48	### Streaming
49
50	```python
51	with client.messages.stream(
52	model="claude-sonnet-4-0",
53	max_tokens=1024,
54	messages=[{"role": "user", "content": "Write a haiku about coding"}]
55	) as stream:
56	for text in stream.text_stream:
57	print(text, end="", flush=True)
58	```
59
60	### System Prompt
61
62	```python
63	message = client.messages.create(
64	model="claude-sonnet-4-0",
65	max_tokens=1024,
66	system="You are a senior Python developer. Be concise.",
67	messages=[{"role": "user", "content": "Review this function"}]
68	)
69	```
70
71	## TypeScript SDK
72
73	### Installation
74
75	```bash
76	npm install @anthropic-ai/sdk
77	```
78
79	### Basic Message
80
81	```typescript
82	import Anthropic from "@anthropic-ai/sdk";
83
84	const client = new Anthropic(); // reads ANTHROPIC_API_KEY from env
85
86	const message = await client.messages.create({
87	model: "claude-sonnet-4-0",
88	max_tokens: 1024,
89	messages: [
90	{ role: "user", content: "Explain async/await in TypeScript" }
91	],
92	});
93	console.log(message.content[0].text);
94	```
95
96	### Streaming
97
98	```typescript
99	const stream = client.messages.stream({
100	model: "claude-sonnet-4-0",
101	max_tokens: 1024,
102	messages: [{ role: "user", content: "Write a haiku" }],
103	});
104
105	for await (const event of stream) {
106	if (event.type === "content_block_delta" && event.delta.type === "text_delta") {
107	process.stdout.write(event.delta.text);
108	}
109	}
110	```
111
112	## Tool Use
113
114	Define tools and let Claude call them:
115
116	```python
117	tools = [
118	{
119	"name": "get_weather",
120	"description": "Get current weather for a location",
121	"input_schema": {
122	"type": "object",
123	"properties": {
124	"location": {"type": "string", "description": "City name"},
125	"unit": {"type": "string", "enum": ["celsius", "fahrenheit"]}
126	},
127	"required": ["location"]
128	}
129	}
130	]
131
132	message = client.messages.create(
133	model="claude-sonnet-4-0",
134	max_tokens=1024,
135	tools=tools,
136	messages=[{"role": "user", "content": "What's the weather in SF?"}]
137	)
138
139	# Handle tool use response
140	for block in message.content:
141	if block.type == "tool_use":
142	# Execute the tool with block.input
143	result = get_weather(**block.input)
144	# Send result back
145	follow_up = client.messages.create(
146	model="claude-sonnet-4-0",
147	max_tokens=1024,
148	tools=tools,
149	messages=[
150	{"role": "user", "content": "What's the weather in SF?"},
151	{"role": "assistant", "content": message.content},
152	{"role": "user", "content": [
153	{"type": "tool_result", "tool_use_id": block.id, "content": str(result)}
154	]}
155	]
156	)
157	```
158
159	## Vision
160
161	Send images for analysis:
162
163	```python
164	import base64
165
166	with open("diagram.png", "rb") as f:
167	image_data = base64.standard_b64encode(f.read()).decode("utf-8")
168
169	message = client.messages.create(
170	model="claude-sonnet-4-0",
171	max_tokens=1024,
172	messages=[{
173	"role": "user",
174	"content": [
175	{"type": "image", "source": {"type": "base64", "media_type": "image/png", "data": image_data}},
176	{"type": "text", "text": "Describe this diagram"}
177	]

People who liked this also viewed

6 picks

Type

claude-api

69stars41forks

View on GitHub

How to install claude-api?

tranhieutt/software_development_department/claude-api

$npx -y skills add tranhieutt/software_development_department --skill claude-api

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Claude API
2
3	Build applications with the Anthropic Claude API and SDKs.
4
5	## When to Activate
6
7	- Building applications that call the Claude API
8	- Code imports `anthropic` (Python) or `@anthropic-ai/sdk` (TypeScript)
9	- User asks about Claude API patterns, tool use, streaming, or vision
10	- Implementing agent workflows with Claude Agent SDK
11	- Optimizing API costs, token usage, or latency
12
13	## Model Selection
14
15	\| Model \| ID \| Best For \|
16	\|-------\|-----\|----------\|
17	\| Opus 4.1 \| `claude-opus-4-1` \| Complex reasoning, architecture, research \|
18	\| Sonnet 4 \| `claude-sonnet-4-0` \| Balanced coding, most development tasks \|
19	\| Haiku 3.5 \| `claude-3-5-haiku-latest` \| Fast responses, high-volume, cost-sensitive \|
20
21	Default to Sonnet 4 unless the task requires deep reasoning (Opus) or speed/cost optimization (Haiku). For production, prefer pinned snapshot IDs over aliases.
22
23	## Python SDK
24
25	### Installation
26
27	```bash
28	pip install anthropic
29	```
30
31	### Basic Message
32
33	```python
34	import anthropic
35
36	client = anthropic.Anthropic() # reads ANTHROPIC_API_KEY from env
37
38	message = client.messages.create(
39	model="claude-sonnet-4-0",
40	max_tokens=1024,
41	messages=[
42	{"role": "user", "content": "Explain async/await in Python"}
43	]
44	)
45	print(message.content[0].text)
46	```
47
48	### Streaming
49
50	```python
51	with client.messages.stream(
52	model="claude-sonnet-4-0",
53	max_tokens=1024,
54	messages=[{"role": "user", "content": "Write a haiku about coding"}]
55	) as stream:
56	for text in stream.text_stream:
57	print(text, end="", flush=True)
58	```
59
60	### System Prompt
61
62	```python
63	message = client.messages.create(
64	model="claude-sonnet-4-0",
65	max_tokens=1024,
66	system="You are a senior Python developer. Be concise.",
67	messages=[{"role": "user", "content": "Review this function"}]
68	)
69	```
70
71	## TypeScript SDK
72
73	### Installation
74
75	```bash
76	npm install @anthropic-ai/sdk
77	```
78
79	### Basic Message
80
81	```typescript
82	import Anthropic from "@anthropic-ai/sdk";
83
84	const client = new Anthropic(); // reads ANTHROPIC_API_KEY from env
85
86	const message = await client.messages.create({
87	model: "claude-sonnet-4-0",
88	max_tokens: 1024,
89	messages: [
90	{ role: "user", content: "Explain async/await in TypeScript" }
91	],
92	});
93	console.log(message.content[0].text);
94	```
95
96	### Streaming
97
98	```typescript
99	const stream = client.messages.stream({
100	model: "claude-sonnet-4-0",
101	max_tokens: 1024,
102	messages: [{ role: "user", content: "Write a haiku" }],
103	});
104
105	for await (const event of stream) {
106	if (event.type === "content_block_delta" && event.delta.type === "text_delta") {
107	process.stdout.write(event.delta.text);
108	}
109	}
110	```
111
112	## Tool Use
113
114	Define tools and let Claude call them:
115
116	```python
117	tools = [
118	{
119	"name": "get_weather",
120	"description": "Get current weather for a location",
121	"input_schema": {
122	"type": "object",
123	"properties": {
124	"location": {"type": "string", "description": "City name"},
125	"unit": {"type": "string", "enum": ["celsius", "fahrenheit"]}
126	},
127	"required": ["location"]
128	}
129	}
130	]
131
132	message = client.messages.create(
133	model="claude-sonnet-4-0",
134	max_tokens=1024,
135	tools=tools,
136	messages=[{"role": "user", "content": "What's the weather in SF?"}]
137	)
138
139	# Handle tool use response
140	for block in message.content:
141	if block.type == "tool_use":
142	# Execute the tool with block.input
143	result = get_weather(**block.input)
144	# Send result back
145	follow_up = client.messages.create(
146	model="claude-sonnet-4-0",
147	max_tokens=1024,
148	tools=tools,
149	messages=[
150	{"role": "user", "content": "What's the weather in SF?"},
151	{"role": "assistant", "content": message.content},
152	{"role": "user", "content": [
153	{"type": "tool_result", "tool_use_id": block.id, "content": str(result)}
154	]}
155	]
156	)
157	```
158
159	## Vision
160
161	Send images for analysis:
162
163	```python
164	import base64
165
166	with open("diagram.png", "rb") as f:
167	image_data = base64.standard_b64encode(f.read()).decode("utf-8")
168
169	message = client.messages.create(
170	model="claude-sonnet-4-0",
171	max_tokens=1024,
172	messages=[{
173	"role": "user",
174	"content": [
175	{"type": "image", "source": {"type": "base64", "media_type": "image/png", "data": image_data}},
176	{"type": "text", "text": "Describe this diagram"}
177	]

People who liked this also viewed

6 picks

Type