dgf1988/codewhale
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

refactor(prompts): decouple allow_shell from static system-prompt prefix

Browse Source 
Move allow_shell from message[0] (static system prompt) to the per-turn
<runtime_prompt> tag alongside mode and approval.  This preserves the
DeepSeek prefix cache across shell-access toggles and mode switches,
which previously forced YOLO entry/exit to rebuild the entire prompt.

Changes:
- Delete remove_shell_tool_guidance and 3 other dead functions (~60 lines)
- Remove allow_shell field from PromptSessionContext and StaticPromptCtx
- Remove shell_tools_available dead parameter from compose functions
- Add Shell Policy section to Runtime Policy Reference (static text)
- Extend <runtime_prompt> tag with allow_shell="true|false" attribute
- Update tests: omits→always_keeps, 83/83 prompts tests pass
- Drop dead compose_mode_prompt_with_approval_and_model

Net: message[0] bytes are now stable regardless of shell-access state.
Mode/approval/shell flags all use the same per-turn tag pattern.
This commit is contained in:
zLeoAlex
2026-06-09 02:36:32 +08:00
parent 9463266cb1
commit 9206eb9d73
3 changed files with 140 additions and 325 deletions
Download Patch File Download Diff File Expand all files Collapse all files
crates/tui/src/core/engine.rs
+12 -11
View File
@@ -681,7 +681,6 @@ impl Engine {
goal_objective_for_prompt(config.goal_objective.as_deref(), &config.goal_state);
let system_prompt =
prompts::system_prompt_for_mode_with_context_skills_session_and_approval(
AppMode::Agent,
&config.workspace,
None,
Some(&config.skills_dir),
@@ -694,7 +693,6 @@ impl Engine {
translation_enabled: config.translation_enabled,
model_id: &config.model,
show_thinking: config.show_thinking,
allow_shell: config.allow_shell,
},
);
let stable_prompt = Some(system_prompt);
@@ -1456,7 +1454,7 @@ impl Engine {
Message {
role: "user".to_string(),
content: vec![ContentBlock::Text {
text: runtime_prompt_text(mode, approval_mode),
text: runtime_prompt_text(mode, approval_mode, self.session.allow_shell),
cache_control: None,
}],
}
@@ -2426,7 +2424,6 @@ impl Engine {
&self.config.goal_state,
);
let base = prompts::system_prompt_for_mode_with_context_skills_session_and_approval(
AppMode::Agent,
&self.config.workspace,
None,
Some(&self.config.skills_dir),
@@ -2439,7 +2436,6 @@ impl Engine {
translation_enabled: self.config.translation_enabled,
model_id: &self.config.model,
show_thinking: self.config.show_thinking,
allow_shell: self.session.allow_shell,
},
);
let mut stable_prompt =
@@ -2649,11 +2645,16 @@ fn agent_approval_mode_for_turn(
/// Produce a minimal runtime-policy tag for the per-turn transient user message.
///
/// All mode and approval policy descriptions live in the frozen system-prompt
/// prefix (`render_runtime_policy_reference()`). This tag is a pointer — the
/// model looks up the corresponding rules from the system prompt. Reduces
/// per-request overhead from ~500 tokens to ~12 tokens.
fn runtime_prompt_text(mode: AppMode, approval_mode: crate::tui::approval::ApprovalMode) -> String {
/// All mode / approval / shell policy descriptions live in the frozen
/// system-prompt prefix (`render_runtime_policy_reference()`). This tag
/// is a pointer — the model looks up the corresponding rules from the
/// system prompt. Keeping these flags out of the static prefix preserves
/// the DeepSeek prefix cache across mode-switches and config-toggles.
fn runtime_prompt_text(
mode: AppMode,
approval_mode: crate::tui::approval::ApprovalMode,
allow_shell: bool,
) -> String {
let mode_str = match mode {
AppMode::Agent => "agent",
AppMode::Plan => "plan",
@@ -2665,7 +2666,7 @@ fn runtime_prompt_text(mode: AppMode, approval_mode: crate::tui::approval::Appro
crate::tui::approval::ApprovalMode::Never => "never",
};
format!(
"<runtime_prompt visibility=\"internal\" mode=\"{mode_str}\" approval=\"{approval_str}\"/>"
"<runtime_prompt visibility=\"internal\" mode=\"{mode_str}\" approval=\"{approval_str}\" allow_shell=\"{allow_shell}\"/>"
)
}
crates/tui/src/prompts.rs
+128 -312
View File
@@ -11,7 +11,6 @@
use crate::models::SystemPrompt;
use crate::project_context::{ProjectContext, load_project_context_with_parents};
use crate::tui::app::AppMode;
use crate::tui::approval::ApprovalMode;
use std::path::{Path, PathBuf};
#[derive(Debug, Clone)]
@@ -39,10 +38,6 @@ pub struct PromptSessionContext<'a> {
/// When false, the prompt should not spend localization pressure on
/// `reasoning_content` the user will never see.
pub show_thinking: bool,
/// Whether shell tools are available in the runtime tool catalog for
/// this session. The prompt must not advertise shell-only workflows
/// when runtime gates have removed those tools.
pub allow_shell: bool,
}
impl Default for PromptSessionContext<'_> {
@@ -55,7 +50,6 @@ impl Default for PromptSessionContext<'_> {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
}
}
}
@@ -314,8 +308,6 @@ pub struct StaticPromptCtx<'a> {
pub model_id: &'a str,
/// Personality overlay requested for the base static prompt.
pub personality: Personality,
/// Whether shell tools are present in the runtime tool catalog.
pub shell_tools_available: bool,
/// Default base/personality prompt layers that would be used without an
/// override.
pub default_layers: &'a str,
@@ -643,6 +635,14 @@ pub const AUTO_APPROVAL: &str = include_str!("prompts/approvals/auto.md");
pub const SUGGEST_APPROVAL: &str = include_str!("prompts/approvals/suggest.md");
pub const NEVER_APPROVAL: &str = include_str!("prompts/approvals/never.md");
/// Shell policy guidance for `allow_shell=false`. Referenced from the
/// Runtime Policy Reference so the model can adapt without mutating the
/// static system-prompt prefix (preserves DeepSeek prefix cache across
/// shell-access toggles).
pub const SHELL_POLICY_DISABLED: &str = "Shell tools unavailable. For mandatory-use items referencing \
`exec_shell`, use `code_execution` (Python sandbox). For GitHub triage, use \
`github_issue_context` / `github_pr_context` as primary route.";
/// Compaction relay template — written into the system prompt so the
/// model knows the format to use when writing `.codewhale/handoff.md`.
pub const COMPACT_TEMPLATE: &str = include_str!("prompts/compact.md");
@@ -758,6 +758,16 @@ pub(crate) fn render_runtime_policy_reference() -> String {
out.push_str("#### never\n\n");
out.push_str(NEVER_APPROVAL.trim());
out.push_str("\n\n");
// ── Shell policy reference ──────────────────────────────────────────
out.push_str("### Shell Policy\n\n");
out.push_str("#### allow_shell=true\n\n");
out.push_str("Shell tools available as described in the base prompt.\n\n");
out.push_str("#### allow_shell=false\n\n");
out.push_str(SHELL_POLICY_DISABLED.trim());
out
}
@@ -843,52 +853,25 @@ directive within Constitutional bounds. Personality, memory, and handoff
context are subordinate to the Constitution, the Statutes, and the user's
current request. When in doubt, consult Article VII: The Hierarchy of Law.";
pub fn compose_prompt(mode: AppMode, personality: Personality) -> String {
compose_prompt_with_approval(mode, personality)
pub fn compose_prompt(personality: Personality) -> String {
compose_prompt_with_approval_model_and_shell(personality, "codewhale")
}
pub fn compose_prompt_with_approval(mode: AppMode, personality: Personality) -> String {
compose_prompt_with_approval_and_model(mode, personality, "codewhale")
}
/// Compose with explicit model ID for dynamic identity injection.
/// The model_id replaces `{model_id}` in the Constitutional preamble.
pub fn compose_prompt_with_approval_and_model(
mode: AppMode,
pub(crate) fn compose_prompt_with_approval_model_and_shell(
personality: Personality,
model_id: &str,
) -> String {
compose_prompt_with_approval_model_and_shell(mode, personality, model_id, true)
}
fn compose_prompt_with_approval_model_and_shell(
mode: AppMode,
personality: Personality,
model_id: &str,
allow_shell: bool,
) -> String {
let shell_tools_available = allow_shell && mode != AppMode::Plan;
let default_layers =
compose_default_static_layers(personality, model_id, shell_tools_available);
let default_layers = compose_default_static_layers(personality, model_id);
apply_static_prompt_composer(
effective_static_prompt_composer(),
personality,
model_id,
shell_tools_available,
&default_layers,
)
}
fn compose_default_static_layers(
personality: Personality,
model_id: &str,
shell_tools_available: bool,
) -> String {
let base_prompt = render_base_prompt_for_tool_availability(
effective_base_prompt().trim(),
model_id,
shell_tools_available,
);
fn compose_default_static_layers(personality: Personality, model_id: &str) -> String {
let base_prompt = apply_model_template(effective_base_prompt().trim(), model_id);
let parts: [&str; 2] = [base_prompt.as_str(), personality.prompt().trim()];
let mut out =
@@ -907,126 +890,31 @@ fn apply_static_prompt_composer(
composer: Option<&StaticPromptComposer>,
personality: Personality,
model_id: &str,
shell_tools_available: bool,
default_layers: &str,
) -> String {
match composer {
Some(composer) => composer(&StaticPromptCtx {
model_id,
personality,
shell_tools_available,
default_layers,
}),
None => default_layers.to_string(),
}
}
fn render_base_prompt_for_tool_availability(
prompt: &str,
model_id: &str,
shell_tools_available: bool,
) -> String {
let prompt = if shell_tools_available {
prompt.to_string()
} else {
remove_shell_tool_guidance(prompt)
};
apply_model_template(&prompt, model_id)
}
fn remove_shell_tool_guidance(prompt: &str) -> String {
let prompt = prompt
.lines()
.filter(|line| !is_shell_disabled_prompt_line(line))
.collect::<Vec<_>>()
.join("\n");
let prompt = remove_markdown_section(&prompt, "### `exec_shell`");
let prompt = prompt.replace(
"; for GitHub issue/PR/release triage, prefer the native `gh ... --json` CLI through shell because it is authenticated, structured, and reproducible; `github_issue_context` / `github_pr_context` are read-only fallbacks when the CLI route is unavailable;",
"; for GitHub issue/PR/release triage, use `github_issue_context` / `github_pr_context` as read-only routes when shell tools are unavailable;",
);
prompt.replace(
"Use deterministic Python inside RLM for exact counts and structured aggregation; use `grep_files` or `exec_shell` directly when that is the clearest deterministic check.",
"Use deterministic Python inside RLM for exact counts and structured aggregation; use `grep_files` directly when that is the clearest deterministic check.",
)
}
fn is_shell_disabled_prompt_line(line: &str) -> bool {
line.starts_with("- Arithmetic, math, calculations → `exec_shell`")
|| line.starts_with("- Hashes, encodings, checksums → `exec_shell`")
|| line.starts_with("- Current time, date, timezone → `exec_shell`")
|| line
.starts_with("- System state: OS, CPU, memory, disk, ports, processes → `exec_shell`")
|| line.starts_with("- **Shell**:")
}
fn remove_markdown_section(prompt: &str, heading: &str) -> String {
let Some(start) = prompt.find(heading) else {
return prompt.to_string();
};
let after_heading = start + heading.len();
let end = prompt[after_heading..]
.find("\n### ")
.map(|offset| after_heading + offset)
.unwrap_or(prompt.len());
let before = prompt[..start].trim_end();
let after = prompt[end..].trim_start_matches('\n');
if before.is_empty() {
after.to_string()
} else if after.is_empty() {
before.to_string()
} else {
format!("{before}\n\n{after}")
}
}
/// Compose for the default personality (Calm).
fn compose_mode_prompt(mode: AppMode) -> String {
compose_prompt(mode, Personality::Calm)
}
fn compose_mode_prompt_with_approval(mode: AppMode) -> String {
compose_prompt_with_approval(mode, Personality::Calm)
}
fn compose_mode_prompt_with_approval_and_model(
mode: AppMode,
_approval_mode: ApprovalMode,
model_id: &str,
) -> String {
compose_prompt_with_approval_model_and_shell(mode, Personality::Calm, model_id, true)
}
// Shell tool guidance removal functions have been deleted.
// The full base prompt is always used; the `allow_shell` flag is
// conveyed via the per-turn <runtime_prompt> tag so the model can
// adapt without mutating the static system-prompt prefix.
// ── Public API ────────────────────────────────────────────────────────
/// Get the system prompt for a specific mode (default Calm personality).
pub fn system_prompt_for_mode(mode: AppMode) -> SystemPrompt {
SystemPrompt::Text(compose_mode_prompt(mode))
}
/// Get the system prompt for a specific mode with explicit personality.
pub fn system_prompt_for_mode_with_personality(
mode: AppMode,
personality: Personality,
) -> SystemPrompt {
SystemPrompt::Text(compose_prompt(mode, personality))
}
/// Get the system prompt for a specific mode with project context.
pub fn system_prompt_for_mode_with_context(
mode: AppMode,
workspace: &Path,
working_set_summary: Option<&str>,
) -> SystemPrompt {
system_prompt_for_mode_with_context_and_skills(
mode,
workspace,
working_set_summary,
None,
None,
None,
)
system_prompt_for_mode_with_context_and_skills(workspace, working_set_summary, None, None, None)
}
/// Get the system prompt for a specific mode with project and skills context.
@@ -1047,7 +935,6 @@ pub fn system_prompt_for_mode_with_context(
/// themselves are turn-volatile. Working-set metadata is now injected into the
/// latest user message as per-turn metadata instead of this system prompt.
pub fn system_prompt_for_mode_with_context_and_skills(
mode: AppMode,
workspace: &Path,
working_set_summary: Option<&str>,
skills_dir: Option<&Path>,
@@ -1055,7 +942,6 @@ pub fn system_prompt_for_mode_with_context_and_skills(
user_memory_block: Option<&str>,
) -> SystemPrompt {
system_prompt_for_mode_with_context_skills_and_session(
mode,
workspace,
working_set_summary,
skills_dir,
@@ -1068,13 +954,11 @@ pub fn system_prompt_for_mode_with_context_and_skills(
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
)
}
pub fn system_prompt_for_mode_with_context_skills_and_session(
mode: AppMode,
workspace: &Path,
_working_set_summary: Option<&str>,
skills_dir: Option<&Path>,
@@ -1082,7 +966,6 @@ pub fn system_prompt_for_mode_with_context_skills_and_session(
session_context: PromptSessionContext<'_>,
) -> SystemPrompt {
system_prompt_for_mode_with_context_skills_session_and_approval(
mode,
workspace,
_working_set_summary,
skills_dir,
@@ -1092,19 +975,14 @@ pub fn system_prompt_for_mode_with_context_skills_and_session(
}
pub fn system_prompt_for_mode_with_context_skills_session_and_approval(
mode: AppMode,
workspace: &Path,
_working_set_summary: Option<&str>,
skills_dir: Option<&Path>,
instructions: Option<&[InstructionSource]>,
session_context: PromptSessionContext<'_>,
) -> SystemPrompt {
let mode_prompt = compose_prompt_with_approval_model_and_shell(
mode,
Personality::Calm,
session_context.model_id,
session_context.allow_shell,
);
let mode_prompt =
compose_prompt_with_approval_model_and_shell(Personality::Calm, session_context.model_id);
// Load project context from workspace
let project_context = load_project_context_with_parents(workspace);
@@ -1176,8 +1054,8 @@ pub fn system_prompt_for_mode_with_context_skills_session_and_approval(
full_prompt = format!("{full_prompt}\n\n{block}");
}
// 4. Context Management (Agent / Yolo only).
if matches!(mode, AppMode::Agent | AppMode::Yolo) {
// 4. Context Management — included in all modes.
{
full_prompt.push_str(
"\n\n## Context Management\n\n\
When the conversation gets long (you'll see a context usage indicator), you can:\n\
@@ -1353,7 +1231,6 @@ mod tests {
let ctx = StaticPromptCtx {
model_id: "deepseek-v4-pro",
personality: Personality::Calm,
shell_tools_available: true,
default_layers: "fallback",
};
@@ -1367,33 +1244,24 @@ mod tests {
#[test]
fn static_prompt_composer_unset_keeps_default_layers_byte_identical() {
for personality in [Personality::Calm, Personality::Playful] {
for shell_tools_available in [true, false] {
let default_layers = compose_default_static_layers(
personality,
"deepseek-v4-flash",
shell_tools_available,
);
let composed = apply_static_prompt_composer(
None,
personality,
"deepseek-v4-flash",
shell_tools_available,
&default_layers,
);
let default_layers = compose_default_static_layers(personality, "deepseek-v4-flash");
let composed = apply_static_prompt_composer(
None,
personality,
"deepseek-v4-flash",
&default_layers,
);
assert_byte_identical("unset static prompt composer", &default_layers, &composed);
}
assert_byte_identical("unset static prompt composer", &default_layers, &composed);
}
}
#[test]
fn static_prompt_composer_receives_context_and_replaces_layers() {
let default_layers =
compose_default_static_layers(Personality::Calm, "deepseek-v4-pro", false);
let default_layers = compose_default_static_layers(Personality::Calm, "deepseek-v4-pro");
let composer: Box<StaticPromptComposer> = Box::new(|ctx| {
assert_eq!(ctx.model_id, "deepseek-v4-pro");
assert_eq!(ctx.personality, Personality::Calm);
assert!(!ctx.shell_tools_available);
assert!(ctx.default_layers.contains("You are deepseek-v4-pro"));
assert!(ctx.default_layers.contains("Personality: Calm"));
assert!(!ctx.default_layers.contains("## Core Tool Taxonomy"));
@@ -1405,7 +1273,6 @@ mod tests {
Some(composer.as_ref()),
Personality::Calm,
"deepseek-v4-pro",
false,
&default_layers,
);
@@ -1506,11 +1373,8 @@ mod tests {
#[test]
fn compose_prompt_injects_model_id() {
let prompt = compose_prompt_with_approval_and_model(
AppMode::Agent,
Personality::Calm,
"deepseek-v4-flash",
);
let prompt =
compose_prompt_with_approval_model_and_shell(Personality::Calm, "deepseek-v4-flash");
assert!(
prompt.contains("You are deepseek-v4-flash"),
"composed prompt must contain the injected model id"
@@ -1522,13 +1386,9 @@ mod tests {
}
#[test]
fn composed_prompt_keeps_shell_guidance_when_shell_tools_are_available() {
let prompt = compose_prompt_with_approval_model_and_shell(
AppMode::Agent,
Personality::Calm,
"deepseek-v4-pro",
true,
);
fn base_prompt_includes_full_shell_tool_guidance() {
let prompt =
compose_prompt_with_approval_model_and_shell(Personality::Calm, "deepseek-v4-pro");
assert!(prompt.contains("- **Shell**:"));
assert!(prompt.contains("### `exec_shell`"));
@@ -1537,15 +1397,15 @@ mod tests {
}
#[test]
fn composed_prompt_omits_shell_guidance_when_shell_tools_are_unavailable() {
let prompt = compose_prompt_with_approval_model_and_shell(
AppMode::Agent,
Personality::Calm,
"deepseek-v4-pro",
false,
);
fn composed_prompt_always_keeps_shell_guidance() {
// After decoupling `allow_shell` from the static system-prompt prefix,
// the base prompt always includes full shell tool guidance. Whether
// shell tools are actually available is conveyed by the per-turn
// <runtime_prompt allow_shell="..."> tag, not by mutating message[0].
let prompt =
compose_prompt_with_approval_model_and_shell(Personality::Calm, "deepseek-v4-pro");
for shell_only in [
for required in [
"- **Shell**:",
"### `exec_shell`",
"`task_shell_start`",
@@ -1555,31 +1415,26 @@ mod tests {
"Hashes, encodings, checksums → `exec_shell`",
"Current time, date, timezone → `exec_shell`",
"System state: OS, CPU, memory, disk, ports, processes → `exec_shell`",
"CLI through shell",
"or `exec_shell` directly",
] {
assert!(
!prompt.contains(shell_only),
"shell-disabled prompt must not advertise {shell_only:?}"
prompt.contains(required),
"static prompt must always include shell guidance: {required:?}"
);
}
assert!(
prompt.contains("actual runtime gates still determine what tools can execute"),
"shell-disabled prompt should keep the runtime-gates hierarchy clause"
"static prompt must include the runtime-gates hierarchy clause"
);
assert!(
prompt.contains("`task_gate_run`") && prompt.contains("`github_issue_context`"),
"shell-disabled prompt should keep non-shell task evidence tools"
"static prompt must include non-shell task evidence tools"
);
}
#[test]
fn composed_prompt_no_longer_inlines_tool_taxonomy() {
let prompt = compose_prompt_with_approval_and_model(
AppMode::Agent,
Personality::Calm,
"deepseek-v4-pro",
);
let prompt =
compose_prompt_with_approval_model_and_shell(Personality::Calm, "deepseek-v4-pro");
// The core tool taxonomy (grep_files / git_status / run_tests hints)
// is no longer prepended as a standalone "## Core Tool Taxonomy" block.
// It now lives inside the "## Runtime Policy Reference" section of the
@@ -1628,7 +1483,6 @@ mod tests {
fn authority_recap_appears_in_full_prompt() {
let tmp = tempdir().expect("tempdir");
let text = match system_prompt_for_mode_with_context_skills_session_and_approval(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -1652,7 +1506,6 @@ mod tests {
fn runtime_policy_reference_is_included_in_full_prompt() {
let tmp = tempdir().expect("tempdir");
let text = match system_prompt_for_mode_with_context_skills_session_and_approval(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -1722,7 +1575,6 @@ mod tests {
write_test_skill(&configured_dir, "configured-skill", "configured skill");
let text = match system_prompt_for_mode_with_context_and_skills(
AppMode::Plan,
&workspace,
None,
Some(&configured_dir),
@@ -1883,7 +1735,6 @@ mod tests {
// both depend on this ordering.
let tmp = tempdir().expect("tempdir");
let text = match system_prompt_for_mode_with_context_skills_session_and_approval(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -1896,7 +1747,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -1954,7 +1804,6 @@ mod tests {
// motivated the closer.
let tmp = tempdir().expect("tempdir");
let text = match system_prompt_for_mode_with_context_skills_session_and_approval(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -1967,7 +1816,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -1998,7 +1846,6 @@ mod tests {
fn hidden_thinking_uses_english_reasoning_without_locale_bookends() {
let tmp = tempdir().expect("tempdir");
let text = match system_prompt_for_mode_with_context_skills_session_and_approval(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -2011,7 +1858,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: false,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -2052,7 +1898,6 @@ mod tests {
// "preamble is opt-in for non-English" invariant.
let tmp = tempdir().expect("tempdir");
let text = match system_prompt_for_mode_with_context_skills_session_and_approval(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -2065,7 +1910,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -2157,7 +2001,6 @@ mod tests {
fn environment_block_is_inserted_into_system_prompt() {
let tmp = tempdir().expect("tempdir");
let prompt = match system_prompt_for_mode_with_context_skills_and_session(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -2170,7 +2013,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -2195,7 +2037,6 @@ mod tests {
fn memory_guidance_absent_when_no_memory_block() {
let tmp = tempdir().expect("tempdir");
let prompt = match system_prompt_for_mode_with_context_skills_and_session(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -2208,7 +2049,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -2225,7 +2065,6 @@ mod tests {
let tmp = tempdir().expect("tempdir");
let block = "## User Memory\n\n- prefers Rust\n";
let prompt = match system_prompt_for_mode_with_context_skills_and_session(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -2238,7 +2077,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -2284,7 +2122,6 @@ mod tests {
let tmp = tempdir().expect("tempdir");
std::fs::write(tmp.path().join("README.md"), "# Pack test").expect("write readme");
let prompt = match system_prompt_for_mode_with_context_skills_and_session(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -2297,7 +2134,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -2314,7 +2150,6 @@ mod tests {
std::fs::write(tmp.path().join(".deepseek").join("handoff.md"), "handoff")
.expect("handoff");
let prompt = match system_prompt_for_mode_with_context_skills_and_session(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -2327,7 +2162,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -2352,7 +2186,7 @@ mod tests {
)
.unwrap();
let prompt = match system_prompt_for_mode_with_context(AppMode::Agent, workspace, None) {
let prompt = match system_prompt_for_mode_with_context(workspace, None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
@@ -2365,7 +2199,7 @@ mod tests {
#[test]
fn missing_handoff_does_not_inject_block() {
let tmp = tempdir().expect("tempdir");
let prompt = match system_prompt_for_mode_with_context(AppMode::Agent, tmp.path(), None) {
let prompt = match system_prompt_for_mode_with_context(tmp.path(), None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
@@ -2378,7 +2212,7 @@ mod tests {
let dir = tmp.path().join(".deepseek");
std::fs::create_dir_all(&dir).unwrap();
std::fs::write(dir.join("handoff.md"), " \n\n ").unwrap();
let prompt = match system_prompt_for_mode_with_context(AppMode::Agent, tmp.path(), None) {
let prompt = match system_prompt_for_mode_with_context(tmp.path(), None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
@@ -2387,7 +2221,7 @@ mod tests {
#[test]
fn compose_prompt_includes_all_layers() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
// Base layer
assert!(prompt.contains("You are codewhale"));
// Personality layer
@@ -2447,7 +2281,7 @@ mod tests {
#[test]
fn compose_prompt_deterministic_order() {
let prompt = compose_prompt(AppMode::Yolo, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
let base_pos = prompt.find("You are codewhale").unwrap();
let personality_pos = prompt.find("Personality: Calm").unwrap();
@@ -2460,23 +2294,19 @@ mod tests {
fn base_prompt_is_mode_agnostic() {
// Mode and approval text are no longer inlined into compose_prompt —
// they travel as request-time runtime metadata.
let agent_prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let yolo_prompt = compose_prompt(AppMode::Yolo, Personality::Calm);
let plan_prompt = compose_prompt(AppMode::Plan, Personality::Calm);
assert!(!agent_prompt.contains("Mode: Agent"));
assert!(!yolo_prompt.contains("Mode: YOLO"));
assert!(!plan_prompt.contains("Mode: Plan"));
assert!(!agent_prompt.contains("Approval Policy:"));
assert!(!yolo_prompt.contains("Approval Policy:"));
assert!(!plan_prompt.contains("Approval Policy:"));
let prompt = compose_prompt(Personality::Calm);
assert!(!prompt.contains("Mode: Agent"));
assert!(!prompt.contains("Mode: YOLO"));
assert!(!prompt.contains("Mode: Plan"));
assert!(!prompt.contains("Approval Policy:"));
// Base prompt still contains Constitutional preamble and personality
assert!(agent_prompt.contains("You are codewhale"));
assert!(agent_prompt.contains("Personality: Calm"));
assert!(prompt.contains("You are codewhale"));
assert!(prompt.contains("Personality: Calm"));
}
#[test]
fn approval_policy_no_longer_inlined_in_base_prompt() {
let prompt = compose_prompt_with_approval(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
assert!(!prompt.contains("Mode: Agent"));
assert!(!prompt.contains("Approval Policy:"));
// Constitutional preamble is still present
@@ -2485,8 +2315,8 @@ mod tests {
#[test]
fn personality_switches_correctly() {
let calm = compose_prompt(AppMode::Agent, Personality::Calm);
let playful = compose_prompt(AppMode::Agent, Personality::Playful);
let calm = compose_prompt(Personality::Calm);
let playful = compose_prompt(Personality::Playful);
assert!(calm.contains("Personality: Calm"));
assert!(playful.contains("Personality: Playful"));
assert!(!calm.contains("Personality: Playful"));
@@ -2495,7 +2325,7 @@ mod tests {
#[test]
fn compact_template_is_included_in_full_prompt() {
let tmp = tempdir().expect("tempdir");
let prompt = match system_prompt_for_mode_with_context(AppMode::Agent, tmp.path(), None) {
let prompt = match system_prompt_for_mode_with_context(tmp.path(), None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
@@ -2516,7 +2346,6 @@ mod tests {
fn session_goal_is_injected_below_compact_template() {
let tmp = tempdir().expect("tempdir");
let prompt = match system_prompt_for_mode_with_context_skills_and_session(
AppMode::Agent,
tmp.path(),
Some("## Repo Working Set\nsrc/lib.rs"),
None,
@@ -2529,7 +2358,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -2552,7 +2380,6 @@ mod tests {
fn empty_session_goal_is_not_injected() {
let tmp = tempdir().expect("tempdir");
let prompt = match system_prompt_for_mode_with_context_skills_and_session(
AppMode::Agent,
tmp.path(),
None,
None,
@@ -2565,7 +2392,6 @@ mod tests {
translation_enabled: false,
model_id: "codewhale",
show_thinking: true,
allow_shell: true,
},
) {
SystemPrompt::Text(text) => text,
@@ -2578,7 +2404,7 @@ mod tests {
#[test]
fn tool_selection_guide_avoids_defensive_tool_suppression() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
assert!(prompt.contains("Tool Selection Guide"));
assert!(prompt.contains("Use `agent_eval`"));
assert!(
@@ -2600,25 +2426,23 @@ mod tests {
/// can't silently weaken the section to a generic "respond in the
/// user's language" directive while keeping the heading.
#[test]
fn language_mirroring_section_present_in_all_modes() {
for mode in [AppMode::Agent, AppMode::Yolo, AppMode::Plan] {
let prompt = compose_prompt(mode, Personality::Calm);
assert!(
prompt.contains("## Language"),
"## Language section missing from mode {mode:?}"
);
assert!(
prompt.contains("reasoning_content"),
"## Language section in {mode:?} must mention `reasoning_content` — \
that field name is the structural anchor for the #588 commitment that \
internal reasoning, not just the visible reply, follows the user's language"
);
}
fn language_mirroring_section_present() {
let prompt = compose_prompt(Personality::Calm);
assert!(
prompt.contains("## Language"),
"## Language section missing from base prompt"
);
assert!(
prompt.contains("reasoning_content"),
"## Language section must mention `reasoning_content` — \
that field name is the structural anchor for the #588 commitment that \
internal reasoning, not just the visible reply, follows the user's language"
);
}
#[test]
fn language_mirroring_prioritizes_latest_user_message_over_locale_default() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
assert!(
prompt.contains("latest user message first"),
"the language directive must choose the turn language from the user message before \
@@ -2644,7 +2468,7 @@ mod tests {
#[test]
fn english_base_prompt_avoids_native_script_language_priming() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
assert!(
!contains_cjk(&prompt),
"English base prompt should keep native-script reinforcement in locale bookends only"
@@ -2663,7 +2487,7 @@ mod tests {
/// changing the wording, don't fail a test for it.
#[test]
fn rlm_specialty_tool_guidance_present() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
// Structural: the RLM heading must exist as a section anchor.
assert!(prompt.contains("RLM — How to Use It"));
// Structural: the word "rlm" must appear multiple times (tool
@@ -2690,7 +2514,7 @@ mod tests {
/// overridable by a single user sentence.
#[test]
fn local_law_tier_covers_engine_config_instructions() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
assert!(
prompt.contains("any file configured via `EngineConfig.instructions`"),
"Tier 5 must explicitly cover EngineConfig.instructions so \
@@ -2700,7 +2524,7 @@ mod tests {
#[test]
fn workspace_orientation_guidance_present() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
assert!(prompt.contains("AGENTS.md"));
assert!(prompt.contains("Local Law"));
assert!(
@@ -2711,7 +2535,7 @@ mod tests {
#[test]
fn prompt_uses_persistent_agent_and_rlm_surface() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
for tool in [
"agent_open",
"agent_eval",
@@ -2749,7 +2573,7 @@ mod tests {
#[test]
fn prompt_documents_fork_context_prefix_cache_contract() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
assert!(prompt.contains("fork_context: true"));
assert!(prompt.contains("byte-identical"));
assert!(prompt.contains("DeepSeek prefix-cache reuse"));
@@ -2758,7 +2582,7 @@ mod tests {
#[test]
fn subagent_done_sentinel_section_present() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
assert!(prompt.contains("Internal Sub-agent Completion Events"));
assert!(prompt.contains("<codewhale:subagent.done>"));
assert!(prompt.contains("not user input"));
@@ -2768,7 +2592,7 @@ mod tests {
#[test]
fn preamble_rhythm_section_present() {
let prompt = compose_prompt(AppMode::Agent, Personality::Calm);
let prompt = compose_prompt(Personality::Calm);
// Preamble rhythm is now part of the Calm personality overlay.
// Verify the load-bearing guidance is still present.
assert!(prompt.contains("In preambles, name the action"));
@@ -2795,16 +2619,14 @@ mod tests {
// Suspect #4 from #263: mode prompt churn within a single mode.
// Two calls with identical (mode, personality) inputs must produce
// identical bytes — anything else is a cache buster.
for mode in [AppMode::Agent, AppMode::Yolo, AppMode::Plan] {
for personality in [Personality::Calm, Personality::Playful] {
let a = compose_prompt(mode, personality);
let b = compose_prompt(mode, personality);
assert_byte_identical(
&format!("compose_prompt(mode={mode:?}, personality={personality:?})"),
&a,
&b,
);
}
for personality in [Personality::Calm, Personality::Playful] {
let a = compose_prompt(personality);
let b = compose_prompt(personality);
assert_byte_identical(
&format!("compose_prompt(personality={personality:?})"),
&a,
&b,
);
}
}
@@ -2822,21 +2644,19 @@ mod tests {
let _skills_dir = EnvVarGuard::remove("DEEPSEEK_SKILLS_DIR");
let workspace = workspace_tmp.path();
for mode in [AppMode::Agent, AppMode::Yolo, AppMode::Plan] {
let a = match system_prompt_for_mode_with_context(mode, workspace, None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
let b = match system_prompt_for_mode_with_context(mode, workspace, None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
assert_byte_identical(
&format!("system_prompt_for_mode_with_context(mode={mode:?}) on empty workspace"),
&a,
&b,
);
}
let a = match system_prompt_for_mode_with_context(workspace, None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
let b = match system_prompt_for_mode_with_context(workspace, None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
assert_byte_identical(
"system_prompt_for_mode_with_context() on empty workspace",
&a,
&b,
);
}
#[test]
@@ -2853,13 +2673,11 @@ mod tests {
let workspace = tmp.path();
let summary = "## Repo Working Set\nWorkspace: /tmp/x\n";
let a = match system_prompt_for_mode_with_context(AppMode::Agent, workspace, Some(summary))
{
let a = match system_prompt_for_mode_with_context(workspace, Some(summary)) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
let b = match system_prompt_for_mode_with_context(AppMode::Agent, workspace, Some(summary))
{
let b = match system_prompt_for_mode_with_context(workspace, Some(summary)) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
@@ -2895,11 +2713,11 @@ mod tests {
)
.unwrap();
let a = match system_prompt_for_mode_with_context(AppMode::Agent, workspace, None) {
let a = match system_prompt_for_mode_with_context(workspace, None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
let b = match system_prompt_for_mode_with_context(AppMode::Agent, workspace, None) {
let b = match system_prompt_for_mode_with_context(workspace, None) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
@@ -2925,11 +2743,10 @@ mod tests {
std::fs::write(handoff_dir.join("handoff.md"), "# handoff body\n").unwrap();
let summary = "## Repo Working Set\nWorkspace: /tmp/x\n";
let prompt =
match system_prompt_for_mode_with_context(AppMode::Agent, workspace, Some(summary)) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
let prompt = match system_prompt_for_mode_with_context(workspace, Some(summary)) {
SystemPrompt::Text(text) => text,
SystemPrompt::Blocks(_) => panic!("expected text system prompt"),
};
let context_pos = prompt
.find("## Context Management")
@@ -3078,7 +2895,6 @@ mod tests {
let extra_source: super::InstructionSource = extra.clone().into();
let prompt = match super::system_prompt_for_mode_with_context_and_skills(
AppMode::Agent,
workspace,
None,
None,
crates/tui/src/tui/ui.rs
-2
View File
@@ -5352,7 +5352,6 @@ async fn dispatch_user_message(
let message_index = app.api_messages.len();
app.system_prompt = Some(
prompts::system_prompt_for_mode_with_context_skills_and_session(
app.mode,
&app.workspace,
None,
None,
@@ -5365,7 +5364,6 @@ async fn dispatch_user_message(
translation_enabled: app.translation_enabled,
model_id: &app.model,
show_thinking: app.show_thinking,
allow_shell: app.allow_shell,
},
),
);