Auto-merge upstream instructkr/claw-code

Author: friuns2

ROADMAP.md

@@ -309,8 +309,9 @@ Priority order: P0 = blocks CI/green state, P1 = blocks integration wiring, P2 =
 20. **Session state classification gap (working vs blocked vs finished vs truly stale)** — **done**: agent manifests now derive machine states such as `working`, `blocked_background_job`, `blocked_merge_conflict`, `degraded_mcp`, `interrupted_transport`, `finished_pending_report`, and `finished_cleanable`, and terminal-state persistence records commit provenance plus derived state so downstream monitoring can distinguish quiet progress from truly idle sessions.
 21. **Resumed `/status` JSON parity gap** — dogfooding shows fresh `claw status --output-format json` now emits structured JSON, but resumed slash-command status still leaks through a text-shaped path in at least one dispatch path. Local CI-equivalent repro fails `rust/crates/rusty-claude-cli/tests/resume_slash_commands.rs::resumed_status_command_emits_structured_json_when_requested` with `expected value at line 1 column 1`, so resumed automation can receive text where JSON was explicitly requested. **Action:** unify fresh vs resumed `/status` rendering through one output-format contract and add regression coverage so resumed JSON output is guaranteed valid.
 22. **Opaque failure surface for session/runtime crashes** — repeated dogfood-facing failures can currently collapse to generic wrappers like `Something went wrong while processing your request. Please try again, or use /new to start a fresh session.` without exposing whether the fault was provider auth, session corruption, slash-command dispatch, render failure, or transport/runtime panic. This blocks fast self-recovery and turns actionable clawability bugs into blind retries. **Action:** preserve a short user-safe failure class (`provider_auth`, `session_load`, `command_dispatch`, `render`, `runtime_panic`, etc.), attach a local trace/session id, and ensure operators can jump from the chat-visible error to the exact failure log quickly.
-23. **`doctor --output-format json` check-level structure gap** — direct dogfooding shows `claw doctor --output-format json` exposes `has_failures` at the top level, but individual check results (`auth`, `config`, `workspace`, `sandbox`, `system`) are buried inside flat prose fields like `message` / `report`. That forces claws to string-scrape human text instead of consuming stable machine-readable diagnostics. **Action:** emit structured per-check JSON (`name`, `status`, `summary`, `details`, and relevant typed fields such as sandbox fallback reason) while preserving the current human-readable report for text mode.
+23. **`doctor --output-format json` check-level structure gap** — **done**: `claw doctor --output-format json` now keeps the human-readable `message`/`report` while also emitting structured per-check diagnostics (`name`, `status`, `summary`, `details`, plus typed fields like workspace paths and sandbox fallback data), with regression coverage in `output_format_contract.rs`.
 24. **Plugin lifecycle init/shutdown test flakes under workspace-parallel execution** — dogfooding surfaced that `build_runtime_runs_plugin_lifecycle_init_and_shutdown` can fail under `cargo test --workspace` while passing in isolation because sibling tests race on tempdir-backed shell init script paths. This is test brittleness rather than a code-path regression, but it still destabilizes CI confidence and wastes diagnosis cycles. **Action:** isolate temp resources per test robustly (unique dirs + no shared cwd assumptions), audit cleanup timing, and add a regression guard so the plugin lifecycle test remains stable under parallel workspace execution.
+26. **Resumed local-command JSON parity gap** — **done**: direct `claw --output-format json` already had structured renderers for `sandbox`, `mcp`, `skills`, `version`, and `init`, but resumed `claw --output-format json --resume <session> /…` paths still fell back to prose because resumed slash dispatch only emitted JSON for `/status`. Resumed `/sandbox`, `/mcp`, `/skills`, `/version`, and `/init` now reuse the same JSON envelopes as their direct CLI counterparts, with regression coverage in `rust/crates/rusty-claude-cli/tests/resume_slash_commands.rs` and `rust/crates/rusty-claude-cli/tests/output_format_contract.rs`.
 **P3 — Swarm efficiency**
 13. Swarm branch-lock protocol — **done**: `branch_lock::detect_branch_lock_collisions()` now detects same-branch/same-scope and nested-module collisions before parallel lanes drift into duplicate implementation
 14. Commit provenance / worktree-aware push events — **done**: lane event provenance now includes branch/worktree/superseded/canonical lineage metadata, and manifest persistence de-dupes superseded commit events before downstream consumers render them

rust/crates/commands/src/lib.rs

@@ -2187,6 +2187,27 @@ pub fn handle_agents_slash_command(args: Option<&str>, cwd: &Path) -> std::io::R
     }
 }
 
+pub fn handle_agents_slash_command_json(args: Option<&str>, cwd: &Path) -> std::io::Result<Value> {
+    if let Some(args) = normalize_optional_args(args) {
+        if let Some(help_path) = help_path_from_args(args) {
+            return Ok(match help_path.as_slice() {
+                [] => render_agents_usage_json(None),
+                _ => render_agents_usage_json(Some(&help_path.join(" "))),
+            });
+        }
+    }
+
+    match normalize_optional_args(args) {
+        None | Some("list") => {
+            let roots = discover_definition_roots(cwd, "agents");
+            let agents = load_agents_from_roots(&roots)?;
+            Ok(render_agents_report_json(cwd, &agents))
+        }
+        Some(args) if is_help_arg(args) => Ok(render_agents_usage_json(None)),
+        Some(args) => Ok(render_agents_usage_json(Some(args))),
+    }
+}
+
 pub fn handle_mcp_slash_command(
     args: Option<&str>,
     cwd: &Path,
@@ -3039,6 +3060,25 @@ fn render_agents_report(agents: &[AgentSummary]) -> String {
     lines.join("\n").trim_end().to_string()
 }
 
+fn render_agents_report_json(cwd: &Path, agents: &[AgentSummary]) -> Value {
+    let active = agents
+        .iter()
+        .filter(|agent| agent.shadowed_by.is_none())
+        .count();
+    json!({
+        "kind": "agents",
+        "action": "list",
+        "working_directory": cwd.display().to_string(),
+        "count": agents.len(),
+        "summary": {
+            "total": agents.len(),
+            "active": active,
+            "shadowed": agents.len().saturating_sub(active),
+        },
+        "agents": agents.iter().map(agent_summary_json).collect::<Vec<_>>(),
+    })
+}
+
 fn agent_detail(agent: &AgentSummary) -> String {
     let mut parts = vec![agent.name.clone()];
     if let Some(description) = &agent.description {
@@ -3327,6 +3367,19 @@ fn render_agents_usage(unexpected: Option<&str>) -> String {
     lines.join("\n")
 }
 
+fn render_agents_usage_json(unexpected: Option<&str>) -> Value {
+    json!({
+        "kind": "agents",
+        "action": "help",
+        "usage": {
+            "slash_command": "/agents [list|help]",
+            "direct_cli": "claw agents [list|help]",
+            "sources": [".claw/agents", "~/.claw/agents", "$CLAW_CONFIG_HOME/agents"],
+        },
+        "unexpected": unexpected,
+    })
+}
+
 fn render_skills_usage(unexpected: Option<&str>) -> String {
     let mut lines = vec![
         "Skills".to_string(),
@@ -3478,6 +3531,18 @@ fn definition_source_json(source: DefinitionSource) -> Value {
     })
 }
 
+fn agent_summary_json(agent: &AgentSummary) -> Value {
+    json!({
+        "name": &agent.name,
+        "description": &agent.description,
+        "model": &agent.model,
+        "reasoning_effort": &agent.reasoning_effort,
+        "source": definition_source_json(agent.source),
+        "active": agent.shadowed_by.is_none(),
+        "shadowed_by": agent.shadowed_by.map(definition_source_json),
+    })
+}
+
 fn skill_origin_id(origin: SkillOrigin) -> &'static str {
     match origin {
         SkillOrigin::SkillsDir => "skills_dir",
@@ -3686,8 +3751,9 @@ pub fn handle_slash_command(
 #[cfg(test)]
 mod tests {
     use super::{
-        handle_plugins_slash_command, handle_skills_slash_command_json, handle_slash_command,
-        load_agents_from_roots, load_skills_from_roots, render_agents_report,
+        handle_agents_slash_command_json, handle_plugins_slash_command,
+        handle_skills_slash_command_json, handle_slash_command, load_agents_from_roots,
+        load_skills_from_roots, render_agents_report, render_agents_report_json,
         render_mcp_report_json_for, render_plugins_report, render_skills_report,
         render_slash_command_help, render_slash_command_help_detail,
         resume_supported_slash_commands, slash_command_specs, suggest_slash_commands,
@@ -4363,6 +4429,72 @@ mod tests {
         let _ = fs::remove_dir_all(user_home);
     }
 
+    #[test]
+    fn renders_agents_reports_as_json() {
+        let workspace = temp_dir("agents-json-workspace");
+        let project_agents = workspace.join(".codex").join("agents");
+        let user_home = temp_dir("agents-json-home");
+        let user_agents = user_home.join(".codex").join("agents");
+
+        write_agent(
+            &project_agents,
+            "planner",
+            "Project planner",
+            "gpt-5.4",
+            "medium",
+        );
+        write_agent(
+            &project_agents,
+            "verifier",
+            "Verification agent",
+            "gpt-5.4-mini",
+            "high",
+        );
+        write_agent(
+            &user_agents,
+            "planner",
+            "User planner",
+            "gpt-5.4-mini",
+            "high",
+        );
+
+        let roots = vec![
+            (DefinitionSource::ProjectCodex, project_agents),
+            (DefinitionSource::UserCodex, user_agents),
+        ];
+        let report = render_agents_report_json(
+            &workspace,
+            &load_agents_from_roots(&roots).expect("agent roots should load"),
+        );
+
+        assert_eq!(report["kind"], "agents");
+        assert_eq!(report["action"], "list");
+        assert_eq!(report["working_directory"], workspace.display().to_string());
+        assert_eq!(report["count"], 3);
+        assert_eq!(report["summary"]["active"], 2);
+        assert_eq!(report["summary"]["shadowed"], 1);
+        assert_eq!(report["agents"][0]["name"], "planner");
+        assert_eq!(report["agents"][0]["model"], "gpt-5.4");
+        assert_eq!(report["agents"][0]["active"], true);
+        assert_eq!(report["agents"][1]["name"], "verifier");
+        assert_eq!(report["agents"][2]["name"], "planner");
+        assert_eq!(report["agents"][2]["active"], false);
+        assert_eq!(report["agents"][2]["shadowed_by"]["id"], "project_claw");
+
+        let help = handle_agents_slash_command_json(Some("help"), &workspace).expect("agents help");
+        assert_eq!(help["kind"], "agents");
+        assert_eq!(help["action"], "help");
+        assert_eq!(help["usage"]["direct_cli"], "claw agents [list|help]");
+
+        let unexpected = handle_agents_slash_command_json(Some("show planner"), &workspace)
+            .expect("agents usage");
+        assert_eq!(unexpected["action"], "help");
+        assert_eq!(unexpected["unexpected"], "show planner");
+
+        let _ = fs::remove_dir_all(workspace);
+        let _ = fs::remove_dir_all(user_home);
+    }
+
     #[test]
     fn lists_skills_from_project_and_user_roots() {
         let workspace = temp_dir("skills-workspace");