@yemi33/minions 0.1.1806 → 0.1.1807

package/CHANGELOG.md

@@ -1,5 +1,10 @@
 # Changelog
 
+## 0.1.1807 (2026-05-08)
+
+### Other
+- docs: clarify detached process handoff
+
 ## 0.1.1806 (2026-05-08)
 
 ### Fixes

package/engine/copilot-models.json

@@ -1,5 +1,5 @@
 {
   "runtime": "copilot",
   "models": null,
-  "cachedAt": "2026-05-08T18:43:02.635Z"
+  "cachedAt": "2026-05-08T19:31:14.792Z"
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@yemi33/minions",
-  "version": "0.1.1806",
+  "version": "0.1.1807",
   "description": "Multi-agent AI dev team that runs from ~/.minions/ — five autonomous agents share a single engine, dashboard, and knowledge base",
   "bin": {
     "minions": "bin/minions.js"

package/playbooks/build-and-test.md

@@ -25,7 +25,7 @@ Work from the current checkout prepared by the engine. Read the repo's own instr
 
 If build or tests fail, report the relevant errors clearly and stop. Do not fix code, push commits, or create PRs from this task.
 
-If a server/app should be run for review, include the URL and a copy-pasteable run command with absolute paths. If the server must survive after the agent exits, start it detached and record the PID, restart command, and stop command; otherwise just provide the command for the user.
+If a server/app should be run for review, include the URL and a copy-pasteable run command with absolute paths. If the server must survive after the agent exits, follow the shared detached-process handoff rules: start it detached with stdio redirected to durable logs, smoke-check it, and record the PID/process identifier, log path, restart command, and stop command; otherwise just provide the command for the user.
 
 ## Findings
 
@@ -35,7 +35,7 @@ Include:
 - Branch, author, and project
 - Build status and important warnings/errors
 - Test status and failed test names if any
-- Local server status, URL, run command, PID, restart command, and stop command if applicable
+- Local server status, URL, run command, PID/process identifier, log path, restart command, and stop command if applicable
 - A short summary of whether the PR is ready to review
 
 ## Constraints

package/playbooks/shared-rules.md

@@ -78,6 +78,13 @@ Use `status: "failed"` plus an accurate `failure_class`, `retryable`, and `needs
 
 Builds, dependency installs, tests, and local servers can be quiet for long periods. Run the repo's normal CLI commands and let them finish; do not add artificial progress output, heartbeat loops, or command-specific workarounds just to keep Minions active.
 
+For finite commands, keep them in the foreground and wait for completion. For persistent dev servers, emulators, file watchers, or other apps that must keep running after you exit, start them as an explicit detached handoff instead of an attached tool command:
+- Detach from the agent shell/session and redirect stdin/stdout/stderr to durable log files. Do not leave a background job inheriting the agent's stdio.
+- Use the OS-native detached process mechanism for the current platform, such as PowerShell `Start-Process -PassThru` on Windows or a POSIX `nohup`/`setsid` launch with `</dev/null` and `>log 2>&1 & echo $!` on macOS/Linux.
+- Capture the PID/process identifier, log path, URL or health-check target, restart command, and stop command with absolute paths.
+- Smoke-check that the URL/emulator/app is responsive before reporting success.
+- State clearly that the detached process is intentionally managed outside the agent/engine lifecycle and must be stopped with the recorded stop command.
+
 ## Done = pushed + local validation. Do NOT wait for remote pipelines.
 
 Your dispatch is **done** the moment (1) your fix is pushed to the branch and (2) any local validation you ran has finished. Write the completion JSON and exit immediately. **Do not** wait for the remote PR pipeline (Android OCM PR build, Espresso CloudTest, GitHub Actions, etc.) to finish before declaring done.

package/playbooks/templates/verify-guide.md

@@ -41,7 +41,9 @@ List anything that requires human judgment:
 ## Manual Testing Guide
 
 **How to run:** (server URL, emulator command, APK path, or N/A)
+**Detached PID / Log:** (PID and durable log path, or N/A)
 **Restart Command:** `cd <absolute-worktree-path> && <command>` (if applicable)
+**Stop Command:** (exact command to stop a detached process, or N/A)
 
 ### <Feature Name> (Plan Item ID)
 **What changed:** brief description

package/playbooks/test.md

@@ -31,18 +31,20 @@ If the task asks you to add or modify files, commit those focused changes, push
 
 If a build or test fails, report the error clearly instead of fixing it unless the task explicitly asks for a fix.
 
-If the task involves a local server or app, report the URL and a ready-to-paste run command with absolute paths:
+If the task involves a local server or app that does not need to stay running after you exit, report the URL and a ready-to-paste run command with absolute paths:
 
 ```text
 ## Run Command
 cd <absolute-path-to-project-or-worktree> && <exact command to start the server>
 ```
 
+If the server/app/emulator must stay running after you exit, follow the shared detached-process handoff rules: start it detached with stdio redirected to durable logs, smoke-check it, and report the PID/process identifier, log path, URL, restart command, and stop command.
+
 ## Findings
 
 Write findings to `{{team_root}}/notes/inbox/{{agent_id}}-{{item_id}}-{{date}}.md` only after successful completion.
 
-Include build status, test results, errors or warnings, run command, localhost URL if applicable, and PR URL if file changes were made.
+Include build status, test results, errors or warnings, run command, localhost URL if applicable, detached PID/log/restart/stop details if applicable, and PR URL if file changes were made.
 
 ## Constraints
 

package/playbooks/verify.md

@@ -25,7 +25,7 @@ Use the setup information in the plan details to create or enter the aggregate w
 
 For each relevant worktree, follow the repo's own instructions (`CLAUDE.md`, README, package files, Makefiles, project scripts), install/restore dependencies as needed, build, and run tests where they exist. If a project fails, report the error and continue verifying the rest; do not turn this into an implementation/fix task.
 
-If there is a user-facing app, identify the normal command and URL. If the app needs to keep running after the agent exits, start it detached in the worktree, save the PID, verify the URL responds, and include restart/stop commands with absolute paths.
+If there is a user-facing app, identify the normal command and URL. If the app needs to keep running after the agent exits, follow the shared detached-process handoff rules: start it detached in the worktree with stdio redirected to durable logs, save the PID/process identifier and log path, verify the URL responds, and include restart/stop commands with absolute paths.
 
 ## Testing Guide