惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

云风的 BLOG
云风的 BLOG
The GitHub Blog
The GitHub Blog
A
About on SuperTechFans
P
Proofpoint News Feed
G
Google Developers Blog
Stack Overflow Blog
Stack Overflow Blog
IT之家
IT之家
Microsoft Security Blog
Microsoft Security Blog
F
Fortinet All Blogs
人人都是产品经理
人人都是产品经理
博客园 - 叶小钗
C
Check Point Blog
Microsoft Azure Blog
Microsoft Azure Blog
aimingoo的专栏
aimingoo的专栏
月光博客
月光博客
美团技术团队
D
Docker
博客园 - Franky
Y
Y Combinator Blog
大猫的无限游戏
大猫的无限游戏
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
博客园 - 【当耐特】
罗磊的独立博客
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报

Recent Commits to openclaw:main

test: merge chat side-result checks · openclaw/openclaw@ddd2c2a test: merge cron history checks · openclaw/openclaw@f7eb746 test: merge responsive navigation shell checks · openclaw/openclaw@c2e4b47 docs(changelog): add codex oauth fixes · openclaw/openclaw@628e6cd test: merge navigation routing cases · openclaw/openclaw@5d8cecb Tests: mock channel registry bundled fallback · openclaw/openclaw@2b08233 Secrets: avoid broad web search discovery for single plugin config · openclaw/openclaw@a464f59 test: merge config view browser checks · openclaw/openclaw@20cf511 fix(status): align oauth health with runtime · openclaw/openclaw@eed7116 feat: add macOS screen snapshots for monitor preview (#67954) thanks … · openclaw/openclaw@f377db1 fix: report shared auth scopes in hello-ok (#67810) thanks @BunsDev · openclaw/openclaw@0b6c39b Auto-reply: avoid eager bundled route fallback · openclaw/openclaw@3ea1bf4 Tests: narrow session binding contract setup · openclaw/openclaw@54e4e16 fix(macOS): enable undo/redo in webchat composer text input (#34962) · openclaw/openclaw@00951dc Tests: speed up channel setup promotion · openclaw/openclaw@82b529a Docs: refresh agent instructions · openclaw/openclaw@5775fe2 fix(auth): serialize OAuth refresh across agents to fix #26322 (#67876) · openclaw/openclaw@8e79080 test: allow ollama public surface boundary test · openclaw/openclaw@7d4f1a6 Docs: add test performance guardrails · openclaw/openclaw@89706d3 Tests: restore context-engine usage proof · openclaw/openclaw@e4c4f95 Tests: slim context engine runtime coverage · openclaw/openclaw@74c198f ci: retry failed custom checkouts · openclaw/openclaw@0ee5baf test: trim duplicate provider auth onboarding cases · openclaw/openclaw@1ffc02e matrix: fix sessions_spawn --thread subagent session spawning (#67643) · openclaw/openclaw@1ce2596 test: reduce auth choice fixture churn · openclaw/openclaw@857b9cd test: mock health status config boundaries · openclaw/openclaw@9d5ab4a test: mock onboard config io boundary · openclaw/openclaw@299694d test: mock legacy state plugin boundaries · openclaw/openclaw@2713089 test: mock channel install boundaries · openclaw/openclaw@b945248 test: mock doctor preview channel boundaries · openclaw/openclaw@b1a3ad4
feat(copilot): add BYOK provider parity · openclaw/opencl...
vincentkoc · 2026-06-24 · via Recent Commits to openclaw:main

@@ -103,8 +103,65 @@ The harness advertises support for the canonical `github-copilot` provider

103103104104

- `github-copilot`

105105106-

Anything outside that set falls through `selection.ts`'s `auto_pi` branch back

107-

to PI.

106+

It also supports custom `models.providers` entries when the selected model has

107+

a non-empty `baseUrl` and one of these API shapes:

108+109+

- `openai-responses`

110+

- `openai-completions`

111+

- `ollama` (OpenAI-compatible completions)

112+

- `azure-openai-responses`

113+

- `anthropic-messages`

114+115+

Native provider ids such as `openai`, `anthropic`, `google`, and `ollama` remain

116+

owned by their native runtimes. Use a distinct custom provider id when routing

117+

an endpoint through Copilot BYOK.

118+119+

Copilot BYOK endpoints must be public-network HTTPS URLs. The harness gives the

120+

Copilot SDK a per-attempt loopback proxy URL, then forwards provider traffic

121+

through OpenClaw's guarded fetch path so DNS pinning and SSRF policy stay

122+

owned by OpenClaw. Use the native OpenClaw runtime for local Ollama, LM Studio,

123+

or LAN model servers.

124+125+

## BYOK

126+127+

Copilot BYOK uses the SDK's session-level custom provider contract. OpenClaw

128+

passes the resolved model endpoint, API key, bearer-token mode, headers, model

129+

id, and context/output limits without moving provider transport logic into

130+

core.

131+132+

For example:

133+134+

```json5

135+

{

136+

agents: {

137+

defaults: {

138+

model: "custom-proxy/llama-3.1-8b",

139+

models: {

140+

"custom-proxy/llama-3.1-8b": {

141+

agentRuntime: { id: "copilot" },

142+

},

143+

},

144+

},

145+

},

146+

models: {

147+

mode: "merge",

148+

providers: {

149+

"custom-proxy": {

150+

baseUrl: "https://api.example.com/v1",

151+

apiKey: "${CUSTOM_PROXY_API_KEY}",

152+

api: "openai-responses",

153+

authHeader: true,

154+

models: [{ id: "llama-3.1-8b", name: "Llama 3.1 8B" }],

155+

},

156+

},

157+

},

158+

}

159+

```

160+161+

BYOK sessions are separately keyed from subscription sessions and from other

162+

endpoints or credential fingerprints. Rotating the key, headers, model, or

163+

endpoint creates a fresh Copilot SDK session instead of resuming incompatible

164+

state.

108165109166

## Auth

110167

@@ -151,10 +208,11 @@ Override with `copilotHome: <path>` on the attempt input when you need a

151208

custom location (for example, a shared mount for migration).

152209153210

Live harness tests use `OPENCLAW_COPILOT_AGENT_LIVE_TOKEN` when a direct token

154-

is needed. The shared live-test setup intentionally scrubs `COPILOT_GITHUB_TOKEN`,

155-

`GH_TOKEN`, and `GITHUB_TOKEN` after staging real auth profiles into the isolated

156-

test home, so passing a `gh auth token` value through the dedicated live-test

157-

variable avoids false skips without exposing the token to unrelated suites.

211+

is needed. The shared live-test setup intentionally scrubs

212+

`COPILOT_GITHUB_TOKEN`, `GH_TOKEN`, and `GITHUB_TOKEN` after staging real auth

213+

profiles into the isolated test home, so passing a `gh auth token` value

214+

through the dedicated live-test variable avoids false skips without exposing

215+

the token to unrelated suites.

158216159217

## Configuration surface

160218

@@ -163,9 +221,9 @@ The harness reads its config from per-attempt input

163221

`extensions/copilot/src/`:

164222165223

- `copilotHome` — per-agent CLI state directory (defaults documented above).

166-

- `model` — string or `{ provider, id, api? }`. When omitted, OpenClaw uses

167-

the agent's normal model selection and the harness verifies the resolved

168-

provider is in the supported set.

224+

- `model` — string or `{ provider, id, api?, baseUrl?, headers?, authHeader? }`.

225+

When omitted, OpenClaw uses the agent's normal model selection and the

226+

harness verifies the resolved provider is supported.

169227

- `reasoningEffort``"low" | "medium" | "high" | "xhigh"`. Maps from

170228

OpenClaw's `ThinkLevel` / `ReasoningLevel` resolution in

171229

`auto-reply/thinking.ts`.

@@ -252,9 +310,9 @@ under `describe("runSideQuestion")`.

252310253311

## Limitations

254312255-

- The harness only claims the canonical `github-copilot` provider at MVP.

256-

Additional providers (BYOK or otherwise) should land in follow-up PRs that

257-

ship the adapter alongside the wire-up.

313+

- The harness claims `github-copilot` plus unowned custom BYOK provider ids.

314+

Manifest-owned native provider ids stay on their owning runtime even when

315+

`agentRuntime.id` is forced to `copilot`.

258316

- The harness does not deliver TUI; PI's TUI is unaffected and remains the

259317

fallback for whatever runtimes do not have a peer surface.

260318

- PI session state is not migrated when an agent switches to `copilot`.