running-claude-code-via-litellm-copilot

Name: running-claude-code-via-litellm-copilot
Author: xixu-me

72stars7forks

How to install running-claude-code-via-litellm-copilot?

xixu-me/skills/running-claude-code-via-litellm-copilot

$npx -y skills add xixu-me/skills --skill running-claude-code-via-litellm-copilot

Installs into the current project.

Summary

Use when routing Claude Code through a local LiteLLM proxy to GitHub Copilot, reducing direct Anthropic spend, configuring ANTHROPIC_BASE_URL or ANTHROPIC_MODEL overrides, or troubleshooting Copilot proxy setup failures such as model-not-found, no localhost traffic, or GitHub 401

Files · 1

View on GitHub

SKILL.md

1	## Overview
2
3	Use this skill for the specific workaround where Claude Code keeps its Anthropic-shaped client behavior, but the actual backend traffic is sent to a local LiteLLM proxy and then forwarded to GitHub Copilot.
4
5	Treat this as an advanced workaround, not an officially guaranteed GitHub workflow. Help the user succeed technically, but do not promise GitHub support, policy approval, or long-term compatibility.
6
7	This skill is guidance-first but execution-aware:
8
9	- If the user wants explanation only, provide the smallest correct set of files, commands, and checks.
10	- If the user wants real setup work on the current machine, inspect first and adapt commands to the active shell and OS.
11	- Pause before persistent edits such as `~/.claude/settings.json` or shell profile files.
12
13	Read [`references/doc-verified-notes.md`](./references/doc-verified-notes.md) before answering if you need to justify which parts come from the article and which parts were tightened against current LiteLLM docs.
14
15	## When To Use
16
17	Use this skill when the user wants any of the following:
18
19	- Claude Code to run against GitHub Copilot through LiteLLM
20	- lower direct Anthropic API spending while keeping the Claude Code workflow
21	- a local `config.yaml` for LiteLLM's GitHub Copilot provider
22	- `ANTHROPIC_BASE_URL`, `ANTHROPIC_MODEL`, or `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` setup
23	- help understanding GitHub device authorization during LiteLLM startup
24	- help with model mismatch, 404-like errors, no requests reaching LiteLLM, or GitHub 401/403 failures
25
26	Do not use this skill for:
27
28	- deciding whether the workaround is allowed by GitHub terms
29	- general LiteLLM architecture unrelated to Claude Code plus Copilot
30	- direct Anthropic API setup with no Copilot or LiteLLM component
31
32	## Core Rules
33
34	1. Lead with a short compliance caveat.
35	Explain that this is a workaround based on a local proxy path, not a GitHub-promoted workflow, and the user must evaluate the latest Copilot terms and limits for themselves.
36	2. Prefer the minimum viable path first.
37	Start with temporary environment variables and a local `config.yaml` unless the user explicitly wants a persistent setup.
38	3. Keep `ANTHROPIC_MODEL` and LiteLLM `model_name` identical.
39	Exact string match matters more than clever explanation.
40	4. Treat `ANTHROPIC_AUTH_TOKEN` as a local placeholder.
41	Claude Code expects a non-empty value locally, but it is not the GitHub Copilot credential and should not be presented as a reusable secret.
42	5. Never overwrite `~/.claude/settings.json` wholesale.
43	Merge only the needed `env` keys and preserve unrelated settings.
44
45	## Workflow
46
47	### 1. Preflight
48
49	Check these first when the user wants real setup work:
50
51	- `claude --help` succeeds
52	- `uv --version` or `pip --version` succeeds
53	- the user has GitHub Copilot access
54	- the intended LiteLLM port is available, usually `4000`
55
56	If the user only wants instructions, state the prerequisites instead of running them.
57
58	### 2. Choose Temporary vs Persistent Setup
59
60	Use this rule:
61
62	- Temporary setup: preferred default for first-time setup, debugging, and low-risk trials
63	- Persistent setup: only when the user explicitly wants the proxy path to apply every time Claude Code starts
64
65	For persistent setup, confirm the target file and then merge keys into `~/.claude/settings.json`. Do not replace the file contents.
66
67	### 3. Create LiteLLM `config.yaml`
68
69	Start from the article's flow, but keep the provider naming aligned with LiteLLM docs:
70
71	```yaml
72	model_list:
73	- model_name: claude-opus-4.5
74	litellm_params:
75	model: github_copilot/claude-opus-4.5
76	drop_params: true
77	```
78
79	Explain the fields:
80
81	- `model_name`: the logical name Claude Code will request
82	- `model`: the LiteLLM provider route, using `github_copilot/<model>`
83	- `drop_params: true`: strips unsupported Anthropic-specific fields before forwarding to Copilot
84
85	If the user wants a different Copilot-backed model, keep the same pattern:
86
87	```yaml
88	model_list:
89	- model_name: <logical-name>
90	litellm_params:
91	model: github_copilot/<copilot-model>
92	drop_params: true
93	```
94
95	Do not hardcode extra headers into the default path unless the user already hit a rejection that suggests header overrides are needed.
96
97	### 4. Install and Start LiteLLM
98
99	Preferred install:
100
101	```bash
102	uv tool install "litellm[proxy]"
103	```
104
105	Fallback:
106
107	```bash
108	pip install "litellm[proxy]"
109	```
110
111	Start the proxy from the directory containing `config.yaml`:
112
113	```bash
114	litellm --config config.yaml --port 4000
115	```
116
117	Tell the user to keep that terminal open because the logs are the fastest truth source during verification.
118
119	### 5. Explain GitHub Device Aut

People who liked this also viewed

6 picks

Type

running-claude-code-via-litellm-copilot

72stars7forks

View on GitHub

How to install running-claude-code-via-litellm-copilot?

xixu-me/skills/running-claude-code-via-litellm-copilot

$npx -y skills add xixu-me/skills --skill running-claude-code-via-litellm-copilot

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	## Overview
2
3	Use this skill for the specific workaround where Claude Code keeps its Anthropic-shaped client behavior, but the actual backend traffic is sent to a local LiteLLM proxy and then forwarded to GitHub Copilot.
4
5	Treat this as an advanced workaround, not an officially guaranteed GitHub workflow. Help the user succeed technically, but do not promise GitHub support, policy approval, or long-term compatibility.
6
7	This skill is guidance-first but execution-aware:
8
9	- If the user wants explanation only, provide the smallest correct set of files, commands, and checks.
10	- If the user wants real setup work on the current machine, inspect first and adapt commands to the active shell and OS.
11	- Pause before persistent edits such as `~/.claude/settings.json` or shell profile files.
12
13	Read [`references/doc-verified-notes.md`](./references/doc-verified-notes.md) before answering if you need to justify which parts come from the article and which parts were tightened against current LiteLLM docs.
14
15	## When To Use
16
17	Use this skill when the user wants any of the following:
18
19	- Claude Code to run against GitHub Copilot through LiteLLM
20	- lower direct Anthropic API spending while keeping the Claude Code workflow
21	- a local `config.yaml` for LiteLLM's GitHub Copilot provider
22	- `ANTHROPIC_BASE_URL`, `ANTHROPIC_MODEL`, or `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` setup
23	- help understanding GitHub device authorization during LiteLLM startup
24	- help with model mismatch, 404-like errors, no requests reaching LiteLLM, or GitHub 401/403 failures
25
26	Do not use this skill for:
27
28	- deciding whether the workaround is allowed by GitHub terms
29	- general LiteLLM architecture unrelated to Claude Code plus Copilot
30	- direct Anthropic API setup with no Copilot or LiteLLM component
31
32	## Core Rules
33
34	1. Lead with a short compliance caveat.
35	Explain that this is a workaround based on a local proxy path, not a GitHub-promoted workflow, and the user must evaluate the latest Copilot terms and limits for themselves.
36	2. Prefer the minimum viable path first.
37	Start with temporary environment variables and a local `config.yaml` unless the user explicitly wants a persistent setup.
38	3. Keep `ANTHROPIC_MODEL` and LiteLLM `model_name` identical.
39	Exact string match matters more than clever explanation.
40	4. Treat `ANTHROPIC_AUTH_TOKEN` as a local placeholder.
41	Claude Code expects a non-empty value locally, but it is not the GitHub Copilot credential and should not be presented as a reusable secret.
42	5. Never overwrite `~/.claude/settings.json` wholesale.
43	Merge only the needed `env` keys and preserve unrelated settings.
44
45	## Workflow
46
47	### 1. Preflight
48
49	Check these first when the user wants real setup work:
50
51	- `claude --help` succeeds
52	- `uv --version` or `pip --version` succeeds
53	- the user has GitHub Copilot access
54	- the intended LiteLLM port is available, usually `4000`
55
56	If the user only wants instructions, state the prerequisites instead of running them.
57
58	### 2. Choose Temporary vs Persistent Setup
59
60	Use this rule:
61
62	- Temporary setup: preferred default for first-time setup, debugging, and low-risk trials
63	- Persistent setup: only when the user explicitly wants the proxy path to apply every time Claude Code starts
64
65	For persistent setup, confirm the target file and then merge keys into `~/.claude/settings.json`. Do not replace the file contents.
66
67	### 3. Create LiteLLM `config.yaml`
68
69	Start from the article's flow, but keep the provider naming aligned with LiteLLM docs:
70
71	```yaml
72	model_list:
73	- model_name: claude-opus-4.5
74	litellm_params:
75	model: github_copilot/claude-opus-4.5
76	drop_params: true
77	```
78
79	Explain the fields:
80
81	- `model_name`: the logical name Claude Code will request
82	- `model`: the LiteLLM provider route, using `github_copilot/<model>`
83	- `drop_params: true`: strips unsupported Anthropic-specific fields before forwarding to Copilot
84
85	If the user wants a different Copilot-backed model, keep the same pattern:
86
87	```yaml
88	model_list:
89	- model_name: <logical-name>
90	litellm_params:
91	model: github_copilot/<copilot-model>
92	drop_params: true
93	```
94
95	Do not hardcode extra headers into the default path unless the user already hit a rejection that suggests header overrides are needed.
96
97	### 4. Install and Start LiteLLM
98
99	Preferred install:
100
101	```bash
102	uv tool install "litellm[proxy]"
103	```
104
105	Fallback:
106
107	```bash
108	pip install "litellm[proxy]"
109	```
110
111	Start the proxy from the directory containing `config.yaml`:
112
113	```bash
114	litellm --config config.yaml --port 4000
115	```
116
117	Tell the user to keep that terminal open because the logs are the fastest truth source during verification.
118
119	### 5. Explain GitHub Device Aut

People who liked this also viewed

6 picks

Type