feat(codex): run context-engine lifecycle in app-server harness (#70809) · openclaw/openclaw@51186d2
jalehman
·
2026-04-24
·
via Recent Commits to openclaw:main
File tree
extensions/codex/src/app-server
| Original file line number | Diff line number | Diff line change |
|---|
@@ -14,6 +14,7 @@ Docs: https://docs.openclaw.ai
|
14 | 14 | - Dependencies/Pi: update bundled Pi packages to `0.70.0`, use Pi's upstream `gpt-5.5` catalog metadata for OpenAI and OpenAI Codex, and keep only local `gpt-5.5-pro` forward-compat handling. |
15 | 15 | - Models/CLI: avoid broad registry enumeration for default `openclaw models list`, reducing default listing latency while preserving configured-row output. (#70883) Thanks @shakkernerd. |
16 | 16 | - Models/CLI: split `openclaw models list` row-source orchestration and registry loading into narrower helpers without changing list output behavior. (#70867) Thanks @shakkernerd. |
| 17 | +- Codex harness/context-engine: run context-engine bootstrap, assembly, post-turn maintenance, and engine-owned compaction in Codex app-server sessions while keeping native Codex thread state and compaction auditable. (#70809) Thanks @jalehman. |
17 | 18 | - Plugins/Google Meet: add a bundled participant plugin with personal Google auth, explicit meeting URL joins, Chrome and Twilio transports, and realtime voice support. (#70765) Thanks @steipete. |
18 | 19 | - Plugins/Google Meet: default Chrome realtime sessions to OpenAI plus SoX `rec`/`play` audio bridge commands, so the usual setup only needs the plugin enabled and `OPENAI_API_KEY`. |
19 | 20 | - Providers/OpenAI: add image generation and reference-image editing through Codex OAuth, so `openai/gpt-image-2` works without an `OPENAI_API_KEY`. Fixes #70703. |
|
| Original file line number | Diff line number | Diff line change |
|---|
@@ -403,6 +403,10 @@
|
403 | 403 | "source": "Tencent Cloud (TokenHub)", |
404 | 404 | "target": "腾讯云（TokenHub）" |
405 | 405 | }, |
| 406 | + { |
| 407 | +"source": "Codex Harness Context Engine Port", |
| 408 | +"target": "Codex Harness Context Engine Port" |
| 409 | + }, |
406 | 410 | { |
407 | 411 | "source": "/gateway/configuration#strict-validation", |
408 | 412 | "target": "/gateway/configuration#strict-validation" |
|
| Original file line number | Diff line number | Diff line change |
|---|
@@ -77,6 +77,10 @@ four lifecycle points:
|
77 | 77 | 4. **After turn** — called after a run completes. The engine can persist state, |
78 | 78 | trigger background compaction, or update indexes. |
79 | 79 | |
| 80 | +For the bundled non-ACP Codex harness, OpenClaw applies the same lifecycle by |
| 81 | +projecting assembled context into Codex developer instructions and the current |
| 82 | +turn prompt. Codex still owns its native thread history and native compactor. |
| 83 | + |
80 | 84 | ### Subagent lifecycle (optional) |
81 | 85 | |
82 | 86 | OpenClaw calls two optional subagent lifecycle hooks: |
|
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。
