wrike

name	wrike
description	Wrike connector skill for managing Agile work items (Epics, Stories, Tasks, sub-tasks, comments, and attachments) via the Wrike REST API v4. Use this skill whenever an agent needs to read or write Wrike workspace data.

Wrike Connector Skill

Connect to a Wrike workspace and manage the full Agile hierarchy: Epics → Stories → Tasks, plus comments and file attachments.

Language Requirement

• The Wrike instance used with this skill is French (FR).
• Any agent using this skill must write titles, descriptions, comments, status notes, and any other content sent to Wrike in French.
• Any communication performed inside Wrike or intended for Wrike updates must be written in French, even if the surrounding task or conversation uses another language.

---

Prerequisites

1. Install dependencies

pip install requests python-dotenv

2. Get a Wrike access token

1. Log in to Wrike
2. Go to Profile → Apps & Integrations → API
3. Click + App, enter a name, click Get Token
4. Copy the token immediately (shown only once)

3. Set environment variables

Export directly:

export WRIKE_ACCESS_TOKEN="your_token_here"
export WRIKE_BASE_URL="https://www.wrike.com/api/v4"   # EU: https://app-eu.wrike.com/api/v4

Or create a .env file next to the scripts/ directory (copy from .env.example):

WRIKE_ACCESS_TOKEN=your_token_here
WRIKE_BASE_URL=https://www.wrike.com/api/v4

4. Verify setup

import sys; sys.path.insert(0, "/path/to/wrike_connector_skill")
from scripts.client import WrikeClient
from scripts import epics

client = WrikeClient()
spaces = epics.get_spaces(client)
print(spaces[0]["title"])   # should print your workspace name

---

How Wrike Maps to Agile Concepts

Agile Concept	Wrike Object	Notes
Epic	Project (folder + metadata)	Has owners, dates, status
Story	Task inside an Epic folder	Has assignees, story points, status
Task / Sub-task	Task with superTaskIds set	Child of a Story
Sprint	Folder inside an Epic	Optional grouping layer
Custom fields	customFields array	Story Points, Acceptance Criteria, etc.
Workflow status	customStatusId on a task	Backlog → In Progress → Review → Done

---

Core Actions

Utility

epics.get_spaces(client)

List all Spaces the user can access. Always call this first to get a space_id.

from scripts import epics
spaces = epics.get_spaces(client)
space_id = spaces[0]["id"]

epics.get_workflows(client)

List all workflows and their custom status IDs. Use to resolve status names to IDs.

workflows = epics.get_workflows(client)
for wf in workflows:
    for status in wf["customStatuses"]:
        print(status["id"], status["name"])

epics.get_custom_fields(client)

List all custom field definitions (IDs, titles, types).

fields = epics.get_custom_fields(client)
sp_field = next(f for f in fields if f["title"] == "Story Points")
sp_field_id = sp_field["id"]

---

Epics

epics.create_epic(client, space_id, title, **kwargs)

Parameter	Type	Required	Description
space_id	str	✅	Space where the Epic is created
title	str	✅	Epic title
description	str		HTML or plain-text description
owner_ids	list[str]		Wrike user IDs of Epic owners
start_date	str		YYYY-MM-DD format
end_date	str		YYYY-MM-DD format
project_status	str		Green/Yellow/Red/Completed/OnHold/Cancelled (default: Green)
custom_fields	list[dict]		[{"id": "...", "value": "..."}]

epic = epics.create_epic(
    client,
    space_id=space_id,
    title="User Authentication",
    description="Everything auth-related",
    start_date="2026-04-01",
    end_date="2026-06-30",
    owner_ids=["KUA..."],
)
epic_id = epic["id"]

epics.get_epic(client, epic_id)

epic = epics.get_epic(client, epic_id)
print(epic["title"], epic["project"]["status"])

epics.list_epics(client, space_id, *, include_archived=False)

Returns only folders that have a project block (actual Projects, not plain folders).

for epic in epics.list_epics(client, space_id):
    print(epic["id"], epic["title"])

epics.find_epic(client, space_id, title)

Find by exact title. Returns None if not found.

epic = epics.find_epic(client, space_id, "User Authentication")

epics.update_epic(client, epic_id, **kwargs)

Same keyword args as create_epic. Only passed fields are changed.

epics.update_epic(client, epic_id, project_status="Completed", end_date="2026-05-01")

epics.delete_epic(client, epic_id)

Moves Epic to Wrike's Recycle Bin.

epics.delete_epic(client, epic_id)

---

Stories

stories.create_story(client, epic_id, title, **kwargs)

Parameter	Type	Required	Description
epic_id	str	✅	Epic (Project) that owns this Story
title	str	✅	Story title
description	str		Description
assignee_ids	list[str]		Wrike user IDs
follower_ids	list[str]		Wrike user IDs
custom_status_id	str		Workflow status ID
importance	str		High/Normal/Low (default: Normal)
start_date	str		YYYY-MM-DD
due_date	str		YYYY-MM-DD
story_points	int		Requires story_points_field_id
story_points_field_id	str		Custom field ID for Story Points
custom_fields	list[dict]		Additional custom fields

story = stories.create_story(
    client,
    epic_id=epic_id,
    title="Login Page",
    assignee_ids=["KUA..."],
    story_points=5,
    story_points_field_id=sp_field_id,
    due_date="2026-04-15",
)
story_id = story["id"]

stories.get_story(client, story_id)

story = stories.get_story(client, story_id)

stories.list_stories(client, epic_id, *, include_completed=True)

Returns direct-child tasks of the Epic (no sub-tasks).

for story in stories.list_stories(client, epic_id, include_completed=False):
    print(story["title"], story["status"])

stories.find_story(client, epic_id, title)

Find by exact title. Returns None if not found.

story = stories.find_story(client, epic_id, "Login Page")

stories.update_story(client, story_id, **kwargs)

Same keyword args as create_story. Only passed fields are changed.

stories.update_story(client, story_id, custom_status_id="IEAD...IN_PROGRESS", importance="High")

stories.delete_story(client, story_id)

stories.delete_story(client, story_id)

---

Tasks (Sub-tasks)

tasks.create_task(client, story_id, epic_id, title, **kwargs)

Parameter	Type	Required	Description
story_id	str	✅	Parent Story (sets superTasks)
epic_id	str	✅	Epic folder (places task in right project)
title	str	✅	Task title
description	str		Description
assignee_ids	list[str]		Wrike user IDs
custom_status_id	str		Workflow status ID
importance	str		High/Normal/Low
start_date	str		YYYY-MM-DD
due_date	str		YYYY-MM-DD
custom_fields	list[dict]		Custom fields

task = tasks.create_task(
    client,
    story_id=story_id,
    epic_id=epic_id,
    title="Write unit tests",
    assignee_ids=["KUA..."],
)

tasks.get_task(client, task_id)

task = tasks.get_task(client, task_id)

tasks.list_tasks(client, story_id, *, include_completed=True)

Returns all sub-tasks of a Story.

for task in tasks.list_tasks(client, story_id):
    print(task["title"])

tasks.update_task(client, task_id, **kwargs)

tasks.update_task(client, task_id, custom_status_id="IEAD...DONE")

tasks.delete_task(client, task_id)

tasks.delete_task(client, task_id)

---

Comments

comments.add_comment(client, task_id, text, *, plain_text=True)

Works on both Stories and Tasks.

comment = comments.add_comment(client, story_id, "Ready for review.")
comment_id = comment["id"]

comments.list_comments(client, task_id, *, plain_text=True)

Returns comments oldest-first.

for c in comments.list_comments(client, story_id):
    print(c["authorId"], c["text"])

comments.update_comment(client, comment_id, text, *, plain_text=True)

comments.update_comment(client, comment_id, "Ready for review — updated.")

comments.delete_comment(client, comment_id)

Permanent — not moved to Recycle Bin.

comments.delete_comment(client, comment_id)

---

Attachments

attachments.upload_attachment(client, task_id, filepath, *, filename=None)

att = attachments.upload_attachment(client, story_id, "/tmp/spec.pdf")
att_id = att["id"]

attachments.list_attachments(client, task_id, *, versions=False)

for a in attachments.list_attachments(client, story_id):
    print(a["name"], a["size"])

attachments.get_attachment(client, attachment_id)

meta = attachments.get_attachment(client, att_id)

attachments.download_attachment(client, attachment_id, dest)

Streams to disk. If dest is a directory the original filename is used.

saved = attachments.download_attachment(client, att_id, "/tmp/downloads/")
print(f"Saved to {saved}")

attachments.delete_attachment(client, attachment_id)

Permanent — not moved to Recycle Bin.

attachments.delete_attachment(client, att_id)

---

Common Workflows

Workflow 1 — Create a full Epic with Stories and Tasks

import sys; sys.path.insert(0, "/path/to/wrike_connector_skill")
from scripts.client import WrikeClient
from scripts import epics, stories, tasks

client = WrikeClient()

# 1. Find your Space
spaces = epics.get_spaces(client)
space_id = spaces[0]["id"]

# 2. Get Story Points field ID
fields = epics.get_custom_fields(client)
sp_id = next((f["id"] for f in fields if f["title"] == "Story Points"), None)

# 3. Create Epic
epic = epics.create_epic(
    client, space_id, "User Authentication",
    start_date="2026-04-01", end_date="2026-06-30",
)

# 4. Create Stories
for title, pts in [("Login Page", 5), ("Password Reset", 3), ("OAuth Integration", 8)]:
    story = stories.create_story(
        client, epic["id"], title,
        story_points=pts, story_points_field_id=sp_id,
    )

# 5. Add sub-tasks to first story
login_story = stories.find_story(client, epic["id"], "Login Page")
for t in ["Design mockup", "Implement form", "Write tests"]:
    tasks.create_task(client, login_story["id"], epic["id"], t)

client.close()

Workflow 2 — Move a Story to "In Progress"

# 1. Resolve the status ID once
workflows = epics.get_workflows(client)
in_progress_id = None
for wf in workflows:
    for s in wf["customStatuses"]:
        if s["name"] == "In Progress":
            in_progress_id = s["id"]
            break

# 2. Find and update the story
story = stories.find_story(client, epic_id, "Login Page")
stories.update_story(client, story["id"], custom_status_id=in_progress_id)

Workflow 3 — Bulk sub-task creation

subtasks = ["Write unit tests", "Code review", "Deploy to staging", "QA sign-off"]
for title in subtasks:
    tasks.create_task(client, story_id, epic_id, title)

Workflow 4 — Attach a file and leave a comment

att = attachments.upload_attachment(client, story_id, "/tmp/design.png")
comments.add_comment(client, story_id, f"Design attached: {att['name']}")

Workflow 5 — Use a CIT template (if configured in workspace)

# CIT must be set up in the Wrike UI first
cit_list = client.get("custom_item_types")
agile_epic_cit = next(c for c in cit_list if c["title"] == "Agile Epic")

result = client.post(
    f"custom_item_types/{agile_epic_cit['id']}/instantiate",
    {"parent": space_id, "title": "New Epic from Template"},
)

---

Decision Rules

Situation	Action
Need to create an Epic	Use epics.create_epic() — creates a Wrike Project
Need to create a Story	Use stories.create_story() — creates a Task inside the Epic folder
Need to create a sub-task	Use tasks.create_task() — sets superTasks to the Story ID
Need workflow status IDs	Call epics.get_workflows() first, resolve by name
Need Story Points field ID	Call epics.get_custom_fields(), filter by title
Got a ResourceNotFoundError	ID is wrong or item was deleted; verify in Wrike UI
Got a RateLimitError	Already retried 3× automatically; reduce call frequency
Got an AuthenticationError	Token is invalid or expired; generate a new one
Workspace is in EU data center	Set WRIKE_BASE_URL=https://app-eu.wrike.com/api/v4
Don't know the Space ID	Call epics.get_spaces(client) and inspect the list
CIT template exists in workspace	Use client.post("custom_item_types/{id}/instantiate", ...) as a shortcut

---

Guardrails & Limitations

• Language: The Wrike workspace language is French (FR). All content created or updated in Wrike must be written in French.
• Rate limit: 400 requests per 60-second window. WrikeClient applies a proactive sliding-window limiter and retries on 429 (up to 3×) before raising RateLimitError.
• Batch GET limit: Wrike allows up to 100 comma-separated IDs per request (e.g. /tasks/id1,id2,...id100).
• ID format: The Wrike UI shows v2 IDs; the API requires v4 IDs. If an ID from the UI returns 404, it may need conversion — copy IDs from API responses, not the browser URL.
• CIT definitions: Cannot be created via API — must be set up in the Wrike UI first.
• Comment/attachment deletion: Permanent (no Recycle Bin). Confirm before calling delete_comment or delete_attachment.
• EU data center: Users on the EU instance must change WRIKE_BASE_URL to https://app-eu.wrike.com/api/v4.
• OAuth tokens: Expire after 1 hour. Permanent tokens never expire. This skill uses permanent tokens by default.
• find_epic / find_story: Do a full list then client-side title match. Avoid in tight loops — cache the result instead.
• Text size limit: When you create the tasks for epics dont add emojis for explenation and be consize and clear in the task body explenation.

---

Environment Variables

Variable	Required	Default	Description
WRIKE_ACCESS_TOKEN	✅	—	Permanent access token
WRIKE_BASE_URL		https://www.wrike.com/api/v4	API base URL (change for EU)

---

File Structure

wrike_connector_skill/
├── wrike/
│   └── SKILL.md              ← this file
├── scripts/
│   ├── __init__.py           ← public re-exports
│   ├── client.py             ← WrikeClient, RateLimiter
│   ├── exceptions.py         ← exception hierarchy
│   ├── epics.py              ← Epic CRUD + utility helpers
│   ├── stories.py            ← Story CRUD
│   ├── tasks.py              ← Task/sub-task CRUD
│   ├── comments.py           ← Comment CRUD
│   └── attachments.py        ← Attachment upload/download/delete
├── tests/
│   ├── test_client.py
│   ├── test_epics.py
│   ├── test_stories.py
│   ├── test_tasks.py
│   └── test_comments.py
├── .env.example
├── requirements.txt
└── AGENTS.md

---

Reference

• Wrike API v4 Documentation
• Wrike Custom Item Types
• Wrike Workflows
• API Status