Skip to content

Worktrunk is a CLI for Git worktree management, designed for parallel AI agent workflows

docs(llm-commits): bump Codex commit-generation model to gpt-5.6-luna (#3430)

Worktrunk Botc532d2dabdaa
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 0e6a9647a..86b55aabd 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -30,6 +30,8 @@
### Documentation
+- **Codex commit-generation model bumped to `gpt-5.6-luna`**: The recommended `[commit.generation]` command for Codex — shown by `wt config create` and in the LLM-commits docs — now uses `gpt-5.6-luna` (was `gpt-5.4-mini`). OpenAI's models page moved `gpt-5.4-mini` into the previous-generation "Other models" tier; `gpt-5.6-luna` is the fast/low-cost variant of the current recommended (5.6) family. ([#3430](https://github.com/max-sixty/worktrunk/pull/3430))
+
- **Code Signing Policy page**: A new page documents Worktrunk's code-signing policy for the Windows binaries under the [SignPath Foundation](https://signpath.org/) open-source program: certificate provenance, the signing pipeline, and per-release approval. Signing addresses Microsoft Defender's false positives on unsigned native binaries. [Docs](https://worktrunk.dev/code-signing/) ([#3366](https://github.com/max-sixty/worktrunk/pull/3366), thanks @bemnlam for reporting [#3355](https://github.com/max-sixty/worktrunk/issues/3355))
### Internal
diff --git a/Taskfile.yaml b/Taskfile.yaml
index 050baf2f9..f055317f3 100644
--- a/Taskfile.yaml
+++ b/Taskfile.yaml
@@ -388,7 +388,7 @@ tasks:
COMMANDS["claude"]='MAX_THINKING_TOKENS=0 claude -p --no-session-persistence --model=haiku --tools='"'"''"'"' --safe-mode --setting-sources='"'"'user'"'"' --system-prompt='"'"''"'"''
COMMANDS["llm"]='llm -m claude-haiku-4.5'
COMMANDS["aichat"]='aichat -m claude:claude-haiku-4.5'
- COMMANDS["codex"]='codex exec -m gpt-5.4-mini -c model_reasoning_effort='"'"'low'"'"' -c system_prompt='"'"''"'"' --sandbox=read-only --json - | jq -sr '"'"'[.[] | select(.item.type? == "agent_message")] | last.item.text'"'"''
+ COMMANDS["codex"]='codex exec -m gpt-5.6-luna -c model_reasoning_effort='"'"'low'"'"' -c system_prompt='"'"''"'"' --sandbox=read-only --json - | jq -sr '"'"'[.[] | select(.item.type? == "agent_message")] | last.item.text'"'"''
COMMANDS["opencode"]='opencode run -m anthropic/claude-haiku-4.5 --variant fast'
# Check which tools are available
diff --git a/dev/config.example.toml b/dev/config.example.toml
index 651b428e2..1a7585a40 100644
--- a/dev/config.example.toml
+++ b/dev/config.example.toml
@@ -67,7 +67,7 @@
# ### Codex
#
# [commit.generation]
-# command = "codex exec -m gpt-5.4-mini -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
+# command = "codex exec -m gpt-5.6-luna -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
#
# ### OpenCode
#
diff --git a/docs/content/config.md b/docs/content/config.md
index 475350cd2..420594d60 100644
--- a/docs/content/config.md
+++ b/docs/content/config.md
@@ -146,7 +146,7 @@ command = "MAX_THINKING_TOKENS=0 claude -p --no-session-persistence --model=haik
```toml
[commit.generation]
-command = "codex exec -m gpt-5.4-mini -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
+command = "codex exec -m gpt-5.6-luna -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
```
### OpenCode
diff --git a/docs/content/llm-commits.md b/docs/content/llm-commits.md
index 598cd8e18..d715cb25c 100644
--- a/docs/content/llm-commits.md
+++ b/docs/content/llm-commits.md
@@ -33,7 +33,7 @@ command = "MAX_THINKING_TOKENS=0 claude -p --no-session-persistence --model=haik
```toml
[commit.generation]
-command = "codex exec -m gpt-5.4-mini -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
+command = "codex exec -m gpt-5.6-luna -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
```
Uses the fast mini model with low reasoning effort and an empty system prompt for faster output. Requires `jq` for JSON parsing. See [Codex CLI docs](https://developers.openai.com/codex/cli/).
diff --git a/skills/worktrunk/reference/config.md b/skills/worktrunk/reference/config.md
index f65c01052..2bfd54204 100644
--- a/skills/worktrunk/reference/config.md
+++ b/skills/worktrunk/reference/config.md
@@ -145,7 +145,7 @@ command = "MAX_THINKING_TOKENS=0 claude -p --no-session-persistence --model=haik
```toml
[commit.generation]
-command = "codex exec -m gpt-5.4-mini -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
+command = "codex exec -m gpt-5.6-luna -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
```
### OpenCode
diff --git a/skills/worktrunk/reference/llm-commits.md b/skills/worktrunk/reference/llm-commits.md
index c69492baa..1a5d6dda0 100644
--- a/skills/worktrunk/reference/llm-commits.md
+++ b/skills/worktrunk/reference/llm-commits.md
@@ -19,7 +19,7 @@ command = "MAX_THINKING_TOKENS=0 claude -p --no-session-persistence --model=haik
```toml
[commit.generation]
-command = "codex exec -m gpt-5.4-mini -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
+command = "codex exec -m gpt-5.6-luna -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
```
Uses the fast mini model with low reasoning effort and an empty system prompt for faster output. Requires `jq` for JSON parsing. See [Codex CLI docs](https://developers.openai.com/codex/cli/).
diff --git a/src/cli/mod.rs b/src/cli/mod.rs
index 36dce2a10..bab0d881b 100644
--- a/src/cli/mod.rs
+++ b/src/cli/mod.rs
@@ -1996,7 +1996,7 @@ command = "MAX_THINKING_TOKENS=0 claude -p --no-session-persistence --model=haik
```toml
[commit.generation]
-command = "codex exec -m gpt-5.4-mini -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
+command = "codex exec -m gpt-5.6-luna -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
```
### OpenCode
diff --git a/src/output/commit_generation.rs b/src/output/commit_generation.rs
index a75676cde..cc5d58188 100644
--- a/src/output/commit_generation.rs
+++ b/src/output/commit_generation.rs
@@ -262,7 +262,7 @@ mod tests {
#[test]
fn test_llm_tool_recommended_config() {
assert_snapshot!(LlmTool::Claude.recommended_config(), @"MAX_THINKING_TOKENS=0 claude -p --no-session-persistence --model=haiku --tools='' --safe-mode --setting-sources='user' --system-prompt=''");
- assert_snapshot!(LlmTool::Codex.recommended_config(), @r#"codex exec -m gpt-5.4-mini -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == "agent_message")] | last.item.text'"#);
+ assert_snapshot!(LlmTool::Codex.recommended_config(), @r#"codex exec -m gpt-5.6-luna -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == "agent_message")] | last.item.text'"#);
assert_snapshot!(LlmTool::OpenCode.recommended_config(), @"opencode run -m anthropic/claude-haiku-4.5 --variant fast");
}
diff --git a/tests/snapshots/integration__integration_tests__help__help_config_create.snap b/tests/snapshots/integration__integration_tests__help__help_config_create.snap
index 4b976119e..5e9b8e856 100644
--- a/tests/snapshots/integration__integration_tests__help__help_config_create.snap
+++ b/tests/snapshots/integration__integration_tests__help__help_config_create.snap
@@ -133,7 +133,7 @@ Creates ~/.config/worktrunk/config.toml with the following content:
  # ### Codex
  #
  # [commit.generation]
-  # command = "codex exec -m gpt-5.4-mini -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
+  # command = "codex exec -m gpt-5.6-luna -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
  #
  # ### OpenCode
  #
diff --git a/tests/snapshots/integration__integration_tests__help__help_config_long.snap b/tests/snapshots/integration__integration_tests__help__help_config_long.snap
index f79874de2..2720590c8 100644
--- a/tests/snapshots/integration__integration_tests__help__help_config_long.snap
+++ b/tests/snapshots/integration__integration_tests__help__help_config_long.snap
@@ -181,7 +181,7 @@ Generate commit messages automatically during merge. Requires an external CLI to
Codex
  [commit.generation]
-  command = "codex exec -m gpt-5.4-mini -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
+  command = "codex exec -m gpt-5.6-luna -c model_reasoning_effort='low' -c system_prompt='' --sandbox=read-only --json - | jq -sr '[.[] | select(.item.type? == \"agent_message\")] | last.item.text'"
OpenCode