Monitoring of Woodpecker CI repositories in Pipedash, with run history, manual pipelines, reruns, cancellation and step logs.
The Woodpecker CI plugin (woodpecker) monitors the latest pipeline of each selected repository. It loads run history, starts manual pipelines on a branch and cancels active runs. It also shows step logs inside Pipedash.
Requirements
The plugin authenticates with a personal access token from the Woodpecker user settings. Starting and cancelling pipelines needs push access to the repository.
Pipedash sends the token in the Authorization: Bearer <token> header and validates it with GET /api/user.
The plugin declares no token prefixes and does not check repository permissions before it offers actions. A successful credential check does not confirm push access.
HTTP 401 reports an authentication error. HTTP 403 reports permission denied, unless rate-limit headers classify it as rate limiting. HTTP 404 reports not found, HTTP 429 a rate limit and HTTP 5xx service unavailable.
Configuration

| Key | Type | Required | Default | Description |
|---|---|---|---|---|
base_url | Text | Yes | None | URL of your Woodpecker server (e.g., https://ci.example.com). |
token | Password | Yes | None | Personal access token from your Woodpecker user settings. Starting and cancelling pipelines needs push access to the repository. |
status_page_url | Text | No | None | Statuspage address for this instance, for example https://status.example.com. Pipedash reads its /api/v2/summary.json. |
The dialog labels these fields "Woodpecker URL", "Personal Access Token" and "Status page URL (optional)".
One more key goes in the config table of the TOML file:
selected_items: Comma-separatedowner/reponames to monitor. Pipedash trims each entry and ignores empty entries. An empty selection shows the errorNo repositories configured.
Base URL
Woodpecker has no hosted default, so base_url is required. The value is the server URL without /api. Pipedash appends /api to every API request.
Pipedash trims surrounding whitespace and trailing slashes. A path prefix stays in place, so https://example.com/woodpecker uses https://example.com/woodpecker/api.
The URL must use http or https and include a host. Query strings and fragments are invalid. On iOS, the URL must use https.
Pipeline discovery
Pipedash reads repositories with one GET /api/user/repos request. The request has no paging parameters, so a server that caps the list also caps discovery. The organization picker lists the distinct owners of those repositories in alphabetical order. It does not call a separate organization endpoint.
The picker shows each repository as owner/repo. Its "Organization" value is the owner and its "Repository" value is the repository name. Entries have no description.
Search matches owner/repo, ignores case and runs before the picker pagination. The organization filter matches the owner exactly. The app loads 100 entries per picker page, and the pagination applies to the returned list.
With one owner, the app selects it automatically. With several owners, the organization is selected before the repositories load.

For each selected repository, Pipedash resolves owner/repo with GET /api/repos/lookup/{owner}/{repo}. It reads the first pipeline with GET /api/repos/{repo_id}/pipelines?page=1&perPage=1.
Each dashboard row represents a repository, not a workflow file or a workflow within a run. Its name is the repository name. Its status and branch come from the latest returned pipeline. A repository with no pipeline shows pending. The row also carries the config_file of the repository as its workflow file.
A failing repository request fails the provider refresh. Pipedash does not skip that repository.
Run history
The app shows 20 runs per history page and sorts by run number, newest first, by default.
The plugin reads GET /api/repos/{repo_id}/pipelines with page and perPage. Each API request asks for at most 50 pipelines. It continues until it reaches the requested history limit or receives an empty page. A short page does not stop the scan, because a server can cap the page size below 50.
Pipedash has no fixed 500-run or 1000-run limit for Woodpecker history. The history view widens its scan when it needs older runs.
Run details read GET /api/repos/{repo_id}/pipelines/{number} directly. They do not depend on finding the run in a bounded history scan.
The run number is the Woodpecker pipeline number. The pipeline ID of the plugin is the numeric repository ID. The run ID has the form woodpecker-run-{id}, using the Woodpecker pipeline id.
Run links open {base_url}/repos/{repo_id}/pipeline/{number}.
Run fields
The history table includes these columns:
| Column | Field | Default visibility |
|---|---|---|
| Run | run_number | Shown |
| Trigger | metadata.event | Shown |
| Status | status | Shown |
| Branch | branch | Shown when present |
| Started | started_at | Shown |
| Duration | duration_seconds | Shown |
| Commit | commit_sha | Hidden, available when present |
| Actor | actor | Hidden, available when present |
The Trigger column shows the event that Woodpecker reports, such as push, pull_request, tag, manual or cron.
Run details carry the branch, author, commit SHA and first line of the commit message. The start time uses started, then created, then the time Pipedash reads the run. The duration is available when both start and finish times exist.
Trigger parameters
| Name | Type | Required | Default | Description |
|---|---|---|---|---|
branch | String | Yes | Default branch of the repository, if present | Branch to run the pipeline on. |
The trigger form labels the field "Branch". Pipedash reads the default branch with GET /api/repos/{repo_id}.
Pipedash starts a manual pipeline with POST /api/repos/{repo_id}/pipelines and the JSON body {"branch":"<branch>"}. It trims the branch before it sends it. A blank branch reports Branch is required to start a Woodpecker pipeline.
The plugin sends no variables, environment overrides, workflow selection or other trigger inputs.
A successful response returns the new pipeline number. If Woodpecker answers 204, Pipedash reports Woodpecker started no pipeline: no workflow runs on manual events.
Rerun
Finished runs offer Rerun. Pipedash starts a new manual pipeline on the branch of the original run. It does not call the restart API of Woodpecker and does not restore the original commit.
If the original run has no branch, the trigger form uses the default branch of the repository when one exists.
Pipedash does not support rerunning only failed jobs or approving a blocked Woodpecker pipeline.
Cancel
Pipedash cancels a running or pending pipeline with POST /api/repos/{repo_id}/pipelines/{number}/cancel. The request has no JSON body.
Cancel uses the pipeline number, not its internal ID. It needs push access to the repository. The plugin offers no force-cancel action.
Run logs
The Logs tab groups steps under their workflows. Each workflow is a group without its own log. Each child step has its own log when its state permits one.
Pipedash reads the workflow and step tree from GET /api/repos/{repo_id}/pipelines/{number}. It reads the log of a step from GET /api/repos/{repo_id}/logs/{number}/{step_id}.
The log endpoint returns JSON entries. Pipedash decodes the base64 data of each entry and adds a newline. An entry without data contributes an empty line.

Log availability
| Step state | Log availability |
|---|---|
running | Partial log with live updates |
pending, blocked, created | Waiting for the log |
skipped, canceled, declined | No log |
| Any other state | Available when the step has a positive start timestamp |
A workflow group has an ID such as workflow:30. A step uses its numeric step ID. The link of a step opens {base_url}/repos/{repo_id}/pipeline/{number}/{pid}.
Live updates
Woodpecker logs update through repeated JSON downloads, not through a streaming connection. Pipedash reads the log of a running step about every 2 seconds while the log is open.
Each request downloads the full entry list. Pipedash remembers the step ID and the number of entries already read, and appends only the new entries to the local log. The cursor counts entries, not bytes, so no range request is involved.
If the entry list becomes shorter, the cursor belongs to another step or the cursor cannot be read, Pipedash replaces the local log.
Pipedash refreshes the workflow and step tree about every 5 seconds while live work remains. A waiting step shows "Logs appear when this job finishes" and opens automatically when its log becomes available.
Log format
Lines that start with ▶ or + show as commands. The first marker has two spaces after the arrow.
The Woodpecker dialect adds no timestamps, sections, error markers or warning markers. Other lines keep the plain marker.
A step-log request has a 15-minute timeout. A missing step log (HTTP 404) reports Logs for step <step_id> are no longer available. Woodpecker has no special handling for HTTP 410. Invalid base64 data reports Invalid Woodpecker log line: <detail>.
The Run logs page covers the viewer.
Status page
Woodpecker has no default status page. The optional status_page_url points Pipedash at a Statuspage address for the instance.
The URL must use https, include a host and contain no credentials, query or fragment. The literal loopback addresses 127.0.0.1 and [::1] can use http. An invalid value reports status_page_url must be an https URL with a host and no credentials, query or fragment.
Pipedash reads api/v2/summary.json below that URL and includes all status-page components. A path prefix stays in place. The plugin adds no separate instance-health checks. See Provider health.
Errors
| Condition | Message |
|---|---|
| Missing or blank URL | Missing Woodpecker base URL |
| Missing or blank token | Missing Woodpecker personal access token |
| No selected repositories | No repositories configured |
| URL with a query or fragment | Invalid Woodpecker URL: remove the query or fragment |
| URL with an unsupported scheme or no host | Woodpecker URL must be an HTTP or HTTPS URL with a host |
| HTTP URL on iOS | Woodpecker on iOS requires an HTTPS server URL, such as https://ci.example.com. Enable HTTPS on your server before adding or importing this provider. |
| Invalid repository or pipeline ID | Invalid pipeline ID format: <id> |
| Missing or blank branch on trigger | Branch is required to start a Woodpecker pipeline |
Manual trigger answers 204 | Woodpecker started no pipeline: no workflow runs on manual events |
| Invalid log step ID | Invalid Woodpecker step ID: <id> |
| Missing step log | Logs for step <step_id> are no longer available |
| Invalid base64 log entry | Invalid Woodpecker log line: <detail> |
TOML example
[providers.woodpecker-home]
name = "Woodpecker Home"
type = "woodpecker"
token = "${WOODPECKER_TOKEN}"
refresh_interval = 30
[providers.woodpecker-home.config]
base_url = "https://ci.example.com"
selected_items = "acme/web,acme/api"
# status_page_url = "https://status.example.com"Status mapping
| Pipedash status | Woodpecker state |
|---|---|
success | success |
failed | failure, error |
running | running |
pending | Any other state, including pending, blocked and created, or a repository with no pipeline |
cancelled | killed, canceled, declined |
skipped | skipped |
The same mapping applies to pipeline runs, workflow groups and steps.