This is the agent-side lifecycle skill. The OpenClaw launcher skill is separate and should be named

preqstation-dispatch

.

Environment

PREQSTATION_API_URL

PREQSTATION API base URL (example:

https://mypreqstation.vercel.app

)

PREQSTATION_TOKEN

PREQSTATION Bearer token (generated in PREQSTATION

/api-keys

)

PREQSTATION_ENGINE

(optional): fallback engine (

claude

|

codex

|

gemini

) when client auto-detection is unavailable

Agent Identity & Engine Mapping

Each agent must identify itself and use the corresponding

engine

value in

all

task operations (list, create, plan, start, update status, complete, review, block):

If you are...

Use

engine=

Claude (Anthropic)

claude

GPT / Codex (OpenAI)

codex

Gemini (Google)

gemini

Always include your

engine

value when listing, creating, planning, starting, completing, reviewing, or blocking tasks.

MCP Plugin Mode (Recommended)

If MCP is available, prefer

scripts/preqstation-mcp-server.mjs

tools.

All mutation tools accept an optional

engine

parameter and always send an engine value using this order:

explicit tool arg

existing task engine (when available)

MCP

initialize.clientInfo.name

auto-detection

PREQSTATION_ENGINE

codex

fallback

Tool

engine usage

preq_list_tasks

Read-only, no engine needed

preq_get_task

Read-only, no engine needed

preq_get_project_settings

Read-only, no engine needed (fetch project deploy settings by key)

preq_plan_task

Assign

engine

when planning task → todo

preq_create_task

Assign

engine

to new inbox task

preq_start_task

Record

engine

starting the work → in_progress

preq_update_task_status

Record

engine

while updating status-only endpoint (

/api/tasks/:id/status

)

preq_complete_task

Record

engine

in work log result → review (fallback to task's existing engine)

preq_review_task

Record

engine

running verification (tests, build, lint) → done

preq_block_task

Record

engine

reporting the block → blocked

preq_delete_task

Permanently delete a task by ticket number or UUID

This gives deterministic task-id based execution and result upload.

Shell Helper Mode

If MCP is not available, source

scripts/preqstation-api.sh

and use shell functions.

All mutation helpers accept an

engine

parameter:

Function

Signature

preq_get_tasks

preq_get_tasks [status] [label]

(read-only, no engine needed)

preq_get_task

preq_get_task

(read-only, no engine needed)

preq_get_project_settings

preq_get_project_settings

(read-only, deploy settings by key)

preq_create_task

preq_create_task ''

(include

engine

in JSON)

preq_patch_task

preq_patch_task ''

(generic PATCH)

preq_start_task

preq_start_task [engine]

preq_update_task_status

preq_update_task_status [engine]

(

status

:

inbox

/

todo

/

in_progress

/

in_review

/

done

/

archived

, supports

review

alias)

preq_plan_task

preq_plan_task [engine]

preq_complete_task

preq_complete_task

[engine] [pr_url] [tests] [notes] [branch_name]

preq_review_task

preq_review_task [engine] [test_cmd] [build_cmd] [lint_cmd]

preq_block_task

preq_block_task [engine]

preq_delete_task

preq_delete_task

preq_resolve_branch_name

preq_resolve_branch_name [fallback_branch]

Requires

jq

for JSON construction in plan/complete/review/block helpers.

For curl examples, see

docs/curl-examples.md

.

Deployment Strategy Contract (required)

Before any git action, resolve deployment strategy through MCP:

Call

preq_get_task

.

Read

task.deploy_strategy

from the response.

If missing, call

preq_get_project_settings

and resolve from:

settings.deploy_strategy

settings.deploy_default_branch

settings.deploy_auto_pr

settings.deploy_commit_on_review

settings.deploy_squash_merge

Expected values from PREQSTATION backend:

strategy

:

direct_commit

|

feature_branch

|

none

default_branch

string (usually

main

)

auto_pr

boolean (feature_branch only)

commit_on_review

boolean

squash_merge

boolean (direct_commit only)

Default when absent/invalid:

strategy=none

default_branch=main

auto_pr=false

commit_on_review=true

squash_merge=true

Behavior by

strategy

:

none

do not run git commit/push/PR. Only code changes + task update result.

direct_commit

merge worktree commits into default_branch and push. No PR. After completing work in the worktree:

In primary checkout: update and merge

git -C < project_cwd

checkout < default_branch

git -C < project_cwd

pull origin < default_branch

squash_merge=true (default): single commit

git -C < project_cwd

merge --squash < worktree_branch

git -C < project_cwd

commit -m ":

"

squash_merge=false: regular merge

git -C < project_cwd

merge < worktree_branch

Push to remote

git

-C

<

project_cwd

>

push origin

<

default_branch

>

feature_branch

push worktree branch to remote. Create PR only when auto_pr=true (requires GitHub MCP on the agent). git -C < project_cwd

push origin < worktree_branch

if auto_pr=true: create PR via GitHub MCP or gh CLI

Rule for commit_on_review : if true and strategy is direct_commit or feature_branch , do not move task to review until remote push is verified. if false , review transition is allowed without mandatory remote push. Execution Flow Call preq_get_task once at the start to fetch task details, acceptance criteria, and the initial status. Resolve the initial status once and execute exactly one matching branch below. Do not chain lifecycle branches in a single run. inbox — plan only: Read local code and prepare the implementation plan. Call preq_plan_task with plan markdown and acceptance criteria. Stop after backend moves the task to todo . Do not implement. todo — start and execute: Call preq_start_task immediately, before code changes or tests, so the task is marked in_progress for the user. Implement code changes and run task-level tests. Resolve deploy strategy via the Deployment Strategy Contract. Perform the required git/deploy steps for direct_commit , feature_branch , or none . Call preq_complete_task with summary, branch, and pr_url when applicable. Stop after backend moves the task to review . Do not call preq_review_task in the same run. in_progress — continue execution: Do not call preq_start_task again. Continue implementation and run task-level tests. Resolve deploy strategy via the Deployment Strategy Contract. Perform the required git/deploy steps for direct_commit , feature_branch , or none . Call preq_complete_task with summary, branch, and pr_url when applicable. Stop after backend moves the task to review . Do not call preq_review_task in the same run. review — verification only: Run verification ( tests , build , lint ). Call preq_review_task on success. Stop after backend moves the task to done . On any failure in an active branch, call preq_block_task with the blocking reason and stop. preq_plan_task , preq_start_task , preq_complete_task , preq_review_task , and preq_block_task are semantic lifecycle actions. Backend owns the actual status transition and must validate the current status for each action. preq_update_task_status is an escape hatch for manual operations, not part of the normal lifecycle flow. preq_complete_task must be used only after the task is in in_progress . preq_review_task must be used only after the task is in review (i.e. after preq_complete_task ). Inbox → Todo Plan Flow User adds short task card to Inbox (optionally specifying engine ). Agent loads candidate tasks with preq_list_tasks (use projectKey + status=inbox + own engine filter). Agent reads local source code and generates implementation plan. Agent calls preq_plan_task with projectKey , taskId , planMarkdown , engine , and optional acceptanceCriteria . Backend uploads plan to task description and moves the card to status=todo .