@@ -79,7 +79,7 @@ Live tests are split into two layers so we can isolate failures:
7979- How to select models:
8080- `OPENCLAW_LIVE_MODELS=modern` to run the modern allowlist (Opus/Sonnet 4.6+, GPT-5.2 + Codex, Gemini 3, DeepSeek V4, GLM 4.7, MiniMax M2.7, Grok 4)
8181- `OPENCLAW_LIVE_MODELS=all` is an alias for the modern allowlist
82-- or `OPENCLAW_LIVE_MODELS="openai/gpt-5.2,openai-codex/gpt-5.2,anthropic/claude-opus-4-6,..."` (comma allowlist)
82+- or `OPENCLAW_LIVE_MODELS="openai/gpt-5.5,openai-codex/gpt-5.5,anthropic/claude-opus-4-6,..."` (comma allowlist)
8383- Modern/all sweeps default to a curated high-signal cap; set `OPENCLAW_LIVE_MAX_MODELS=0` for an exhaustive modern sweep or a positive number for a smaller cap.
8484- Exhaustive sweeps use `OPENCLAW_LIVE_TEST_TIMEOUT_MS` for the whole direct-model test timeout. Default: 60 minutes.
8585- Direct-model probes run with 20-way parallelism by default; set `OPENCLAW_LIVE_MODEL_CONCURRENCY` to override.
@@ -149,7 +149,7 @@ openclaw models list --json
149149- Default provider/model: `claude-cli/claude-sonnet-4-6`
150150- Command/args/image behavior come from the owning CLI backend plugin metadata.
151151- Overrides (optional):
152-- `OPENCLAW_LIVE_CLI_BACKEND_MODEL="codex-cli/gpt-5.2"`
152+- `OPENCLAW_LIVE_CLI_BACKEND_MODEL="codex-cli/gpt-5.5"`
153153- `OPENCLAW_LIVE_CLI_BACKEND_COMMAND="/full/path/to/codex"`
154154- `OPENCLAW_LIVE_CLI_BACKEND_ARGS='["exec","--json","--color","never","--sandbox","read-only","--skip-git-repo-check"]'`
155155- `OPENCLAW_LIVE_CLI_BACKEND_IMAGE_PROBE=1` to send a real image attachment (paths are injected into the prompt). Docker recipes default this off unless explicitly requested.
@@ -163,7 +163,7 @@ Example:
163163164164```bash
165165OPENCLAW_LIVE_CLI_BACKEND=1 \
166- OPENCLAW_LIVE_CLI_BACKEND_MODEL="codex-cli/gpt-5.2" \
166+ OPENCLAW_LIVE_CLI_BACKEND_MODEL="codex-cli/gpt-5.5" \
167167 pnpm test:live src/gateway/gateway-cli-backend.live.test.ts
168168```
169169@@ -227,11 +227,11 @@ Notes:
227227- `OPENCLAW_LIVE_ACP_BIND_AGENT=opencode`
228228- `OPENCLAW_LIVE_ACP_BIND_AGENTS=claude,codex,gemini`
229229- `OPENCLAW_LIVE_ACP_BIND_AGENT_COMMAND='npx -y @agentclientprotocol/claude-agent-acp@<version>'`
230-- `OPENCLAW_LIVE_ACP_BIND_CODEX_MODEL=gpt-5.2`
230+- `OPENCLAW_LIVE_ACP_BIND_CODEX_MODEL=gpt-5.5`
231231- `OPENCLAW_LIVE_ACP_BIND_OPENCODE_MODEL=opencode/kimi-k2.6`
232232- `OPENCLAW_LIVE_ACP_BIND_REQUIRE_TRANSCRIPT=1`
233233- `OPENCLAW_LIVE_ACP_BIND_REQUIRE_CRON=1`
234-- `OPENCLAW_LIVE_ACP_BIND_PARENT_MODEL=openai/gpt-5.2`
234+- `OPENCLAW_LIVE_ACP_BIND_PARENT_MODEL=openai/gpt-5.5`
235235- Notes:
236236- This lane uses the gateway `chat.send` surface with admin-only synthetic originating-route fields so tests can attach message-channel context without pretending to deliver externally.
237237- When `OPENCLAW_LIVE_ACP_BIND_AGENT_COMMAND` is unset, the test uses the embedded `acpx` plugin's built-in agent registry for the selected ACP harness agent.
@@ -277,7 +277,7 @@ Docker notes:
277277`agent` method:
278278- load the bundled `codex` plugin
279279- select `OPENCLAW_AGENT_RUNTIME=codex`
280-- send a first gateway agent turn to `openai/gpt-5.2` with the Codex harness forced
280+- send a first gateway agent turn to `openai/gpt-5.5` with the Codex harness forced
281281- send a second turn to the same OpenClaw session and verify the app-server
282282 thread can resume
283283- run `/codex status` and `/codex models` through the same gateway command
@@ -287,7 +287,7 @@ Docker notes:
287287 denied so the agent asks back
288288- Test: `src/gateway/gateway-codex-harness.live.test.ts`
289289- Enable: `OPENCLAW_LIVE_CODEX_HARNESS=1`
290-- Default model: `openai/gpt-5.2`
290+- Default model: `openai/gpt-5.5`
291291- Optional image probe: `OPENCLAW_LIVE_CODEX_HARNESS_IMAGE_PROBE=1`
292292- Optional MCP/tool probe: `OPENCLAW_LIVE_CODEX_HARNESS_MCP_PROBE=1`
293293- Optional Guardian probe: `OPENCLAW_LIVE_CODEX_HARNESS_GUARDIAN_PROBE=1`
@@ -305,7 +305,7 @@ OPENCLAW_LIVE_CODEX_HARNESS=1 \
305305 OPENCLAW_LIVE_CODEX_HARNESS_IMAGE_PROBE=1 \
306306 OPENCLAW_LIVE_CODEX_HARNESS_MCP_PROBE=1 \
307307 OPENCLAW_LIVE_CODEX_HARNESS_GUARDIAN_PROBE=1 \
308- OPENCLAW_LIVE_CODEX_HARNESS_MODEL=openai/gpt-5.2 \
308+ OPENCLAW_LIVE_CODEX_HARNESS_MODEL=openai/gpt-5.5 \
309309 pnpm test:live -- src/gateway/gateway-codex-harness.live.test.ts
310310```
311311@@ -336,13 +336,13 @@ Docker notes:
336336Narrow, explicit allowlists are fastest and least flaky:
337337338338- Single model, direct (no gateway):
339-- `OPENCLAW_LIVE_MODELS="openai/gpt-5.2" pnpm test:live src/agents/models.profiles.live.test.ts`
339+- `OPENCLAW_LIVE_MODELS="openai/gpt-5.5" pnpm test:live src/agents/models.profiles.live.test.ts`
340340341341- Single model, gateway smoke:
342-- `OPENCLAW_LIVE_GATEWAY_MODELS="openai/gpt-5.2" pnpm test:live src/gateway/gateway-models.profiles.live.test.ts`
342+- `OPENCLAW_LIVE_GATEWAY_MODELS="openai/gpt-5.5" pnpm test:live src/gateway/gateway-models.profiles.live.test.ts`
343343344344- Tool calling across several providers:
345-- `OPENCLAW_LIVE_GATEWAY_MODELS="openai/gpt-5.2,openai-codex/gpt-5.2,anthropic/claude-opus-4-6,google/gemini-3-flash-preview,deepseek/deepseek-v4-flash,zai/glm-5.1,minimax/MiniMax-M2.7" pnpm test:live src/gateway/gateway-models.profiles.live.test.ts`
345+- `OPENCLAW_LIVE_GATEWAY_MODELS="openai/gpt-5.5,openai-codex/gpt-5.5,anthropic/claude-opus-4-6,google/gemini-3-flash-preview,deepseek/deepseek-v4-flash,zai/glm-5.1,minimax/MiniMax-M2.7" pnpm test:live src/gateway/gateway-models.profiles.live.test.ts`
346346347347- Google focus (Gemini API key + Antigravity):
348348- Gemini (API key): `OPENCLAW_LIVE_GATEWAY_MODELS="google/gemini-3-flash-preview" pnpm test:live src/gateway/gateway-models.profiles.live.test.ts`
@@ -370,8 +370,8 @@ There is no fixed “CI model list” (live is opt-in), but these are the **reco
370370371371This is the “common models” run we expect to keep working:
372372373-- OpenAI (non-Codex): `openai/gpt-5.2`
374-- OpenAI Codex OAuth: `openai-codex/gpt-5.2`
373+- OpenAI (non-Codex): `openai/gpt-5.5`
374+- OpenAI Codex OAuth: `openai-codex/gpt-5.5`
375375- Anthropic: `anthropic/claude-opus-4-6` (or `anthropic/claude-sonnet-4-6`)
376376- Google (Gemini API): `google/gemini-3.1-pro-preview` and `google/gemini-3-flash-preview` (avoid older Gemini 2.x models)
377377- Google (Antigravity): `google-antigravity/claude-opus-4-6-thinking` and `google-antigravity/gemini-3-flash`
@@ -380,13 +380,13 @@ This is the “common models” run we expect to keep working:
380380- MiniMax: `minimax/MiniMax-M2.7`
381381382382Run gateway smoke with tools + image:
383-`OPENCLAW_LIVE_GATEWAY_MODELS="openai/gpt-5.2,openai-codex/gpt-5.2,anthropic/claude-opus-4-6,google/gemini-3.1-pro-preview,google/gemini-3-flash-preview,google-antigravity/claude-opus-4-6-thinking,google-antigravity/gemini-3-flash,deepseek/deepseek-v4-flash,zai/glm-5.1,minimax/MiniMax-M2.7" pnpm test:live src/gateway/gateway-models.profiles.live.test.ts`
383+`OPENCLAW_LIVE_GATEWAY_MODELS="openai/gpt-5.5,openai-codex/gpt-5.5,anthropic/claude-opus-4-6,google/gemini-3.1-pro-preview,google/gemini-3-flash-preview,google-antigravity/claude-opus-4-6-thinking,google-antigravity/gemini-3-flash,deepseek/deepseek-v4-flash,zai/glm-5.1,minimax/MiniMax-M2.7" pnpm test:live src/gateway/gateway-models.profiles.live.test.ts`
384384385385### Baseline: tool calling (Read + optional Exec)
386386387387Pick at least one per provider family:
388388389-- OpenAI: `openai/gpt-5.2`
389+- OpenAI: `openai/gpt-5.5`
390390- Anthropic: `anthropic/claude-opus-4-6` (or `anthropic/claude-sonnet-4-6`)
391391- Google: `google/gemini-3-flash-preview` (or `google/gemini-3.1-pro-preview`)
392392- DeepSeek: `deepseek/deepseek-v4-flash`
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。