Path to Prod

by Alexandre Charbit
5
4
3
2
1
Score: 50/100

Description

Forecast Jira delivery milestones from workflow history, calendars, and team capacity.

Reviews

No reviews yet.

Stats

0
stars
36
downloads
0
forks
3
days
1
days
1
days
0
total PRs
0
open PRs
0
closed PRs
0
merged PRs
0
total issues
0
open issues
0
closed issues
9
commits

Latest Version

2 days ago

Changelog

What's new

  • Connect to Jira Data Center/Server with its REST API v2, Bearer personal access tokens, company-hosted Jira URLs, and optional proxy base paths such as /jira.
  • Forecast Data Center epics through the Epic Link field and read issue status history from expanded changelogs. Cloud REST API v3 remains supported.
  • Test the configured project's issue search and changelog access directly. Proxy authentication and HTML SSO responses now produce clearer errors without displaying response contents.
  • Store Jira tokens in Obsidian SecretStorage. Existing tokens in plugin settings migrate on load; only the secret name remains in plugin data.
  • Read epic timelines more easily with distinct P80 milestone markers, a today guide, and clearer environment date cards. Private calendar feed URLs are no longer included in connection errors.

Upgrade notes

Choose Data center / server in Path to Prod settings if automatic detection does not match your company's Jira URL. Leave the email field blank for a Data Center personal access token. Obsidian 1.13.0 or later is required. Older backups of the plugin's data.json may still contain a token from previous releases; handle those copies under your company's credential policy.

If Jira truncates an expanded Data Center changelog, Path to Prod stops instead of calculating from incomplete history. The Data Center path has automated tests but has not been verified against your company's live Jira instance.

README file from

Github

Path to Prod

Path to Prod is an Obsidian plugin that forecasts when a Jira goal will finish development, staging, and production. It does not require story points. Instead, it learns stage durations from completed Jira issues, finds comparable tickets, and runs a capacity-aware Monte Carlo simulation.

Features

  • Reads goal and historical tickets with configurable JQL.
  • Reconstructs stage durations from Jira status changelogs.
  • Matches unpointed tickets using issue type, component, labels, priority, and summary words.
  • Produces P50, P80, and P95 milestone dates.
  • Forecasts Jira epics from their child tickets, with an editable reserve for unrefined work, and shows development, staging, and production on a Gantt timeline.
  • Shows each environment's P80 date beneath its colored bar and a dashed range from production P80 to P95.
  • Labels the epic timeline by ISO week number and Monday start date, including the correct ISO week year around New Year.
  • Copies the current dashboard as an email-ready report, with colored stage cells where rich paste is supported and a complete plain-text fallback.
  • Watches important epics and tickets selected in the dashboard or identified by configurable Jira priorities and labels. Jira due dates take precedence; Scrum work without due dates is compared with a configurable sprint target.
  • Auto-detects Scrum versus Kanban from Jira sprint metadata, with a manual override.
  • Aligns Scrum development to configurable 1–6 week sprint boundaries; Kanban stays continuous-flow and WIP-capacity based.
  • Accounts for weekdays, manual or ICS-synchronized holidays, crew leave, focus factor, parallel capacity, and production release days.
  • Stores the Jira token in Obsidian SecretStorage and keeps forecast history in the vault's local plugin data.

Forecast dates are probabilistic planning aids, not commitments. Sparse history, workflow changes, unrecorded dependencies, and inaccurate calendar data can materially affect results.

Development

npm install
npm run check

To test in Obsidian, copy or symlink this directory into a development vault at .obsidian/plugins/path-to-prod, run npm run dev, reload Obsidian, and enable Path to Prod under Community plugins. Use a separate development vault.

Configure

  1. Create a dedicated Jira credential: a Jira Data Center/Server personal access token, an Atlassian Cloud API token, or another company-approved Bearer token.
  2. Enter the company-approved HTTPS Jira API URL (not a project or board page), token holder, and company Jira project key or exact name. Select or create an Obsidian secret for the token. Auto detect uses Data Center/Server REST API v2 for company-hosted URLs and Cloud REST API v3 for Atlassian hosts; select the deployment manually if your proxy uses a different hostname. Leave email blank for Data Center/Server personal access tokens or other Bearer authentication. Atlassian Cloud API tokens using Basic authentication require the Atlassian account email. Scoped Atlassian tokens use https://api.atlassian.com/ex/jira/<cloudId> as the API base URL.
  3. Use Test Jira connection to run project issue search and, if an issue is found, read its changelog as forecasting requires. A company proxy path such as /jira is supported in the base URL. A 407 error points to proxy authentication; HTML instead of JSON usually means a proxy, SSO, or login page is intercepting the API call. Ask your administrator to allow Obsidian access to the approved Jira API host and trust the required company TLS certificate. Do not put proxy credentials in the Jira token field. Data Center returns changelogs through expand=changelog; if the server truncates them, Path to Prod stops rather than forecasting from incomplete history.
  4. Set a completed-ticket JQL query and a goal JQL query. Use {{project}} to reference the configured company project key.
  5. Map the team's Jira statuses to development, staging, and production.
  6. Select Scrum, Kanban, or auto detection. For Scrum, set the sprint length and a known sprint start date.
  7. Add holidays, leave, capacity, focus factor, and release weekdays. Optionally add HTTPS ICS holiday feeds and one all-day leave feed per crew member.
  8. Open Path to Prod, sync history, and run the forecast.

To forecast epics, set Epic JQL and open Epic timeline. Adjust the unrefined reserve for each epic as its scope becomes clearer. Set sprint length and the important-work target in the dashboard. Jira priorities and labels can mark work as important automatically, and each epic or child ticket can also be marked manually. The watch compares the epic's P80 production date with the target for its important scope; it is a signal to review the plan, not an individual ticket completion forecast. Kanban work without a Jira due date is shown as needing a target rather than given a sprint alert. Each epic uses the full configured team capacity independently, so simultaneous epic dates are not a combined portfolio commitment.

Use Email draft on the goal or epic dashboard to open the default email client with a suggested subject and copy the current forecast for pasting into the draft. Copy for email only copies the report. Rich paste retains the colored environment cells; clients that only support plain text receive the same dates, ISO week numbers, and important-work alerts as text. The plugin does not send email or choose recipients.

Path to Prod supports Basic authentication for Atlassian Cloud API tokens and Bearer authentication when the email field is blank, including Jira Data Center/Server personal access tokens. The Jira token is stored in Obsidian SecretStorage; the plugin data keeps only its secret name. Existing plaintext tokens are migrated when the plugin loads. Use a revocable token and protect access to the vault.

Support

Path to Prod is free. If you would like to support ongoing development, you can do so optionally on Ko-fi. Donations do not unlock features or change forecast behavior.

Privacy and network access

Path to Prod initiates network requests to the configured Jira API host and optional HTTPS ICS calendar feeds. Obsidian's requestUrl handles these requests, so the plugin has no separate proxy service or proxy credential setting. Corporate proxy, VPN, certificate, and allowlist policy still apply to the app's network connection. The Jira token is stored in Obsidian SecretStorage; the token-holder label, project key, private calendar URLs, cached issue metadata, and crew leave dates are stored in the vault's local plugin data file. The plugin has no telemetry, analytics, advertising, or developer-operated server. Protect the vault and use dedicated, revocable credentials with the least Jira permissions that allow the required project and issues to be read. Do not sync the plugin data file to an untrusted location. Older copies or backups of data.json may still contain a token saved by earlier releases; remove or rotate those separately according to company policy.

Forecast model

For every simulation, the engine:

  1. Finds up to 20 comparable historical issues for each goal ticket and stage.
  2. Samples a duration from the comparable distribution.
  3. Falls back to configurable assumptions when a stage lacks history.
  4. Converts total stage effort into dates using team capacity and calendar availability.
  5. Snaps the production milestone to an allowed release weekday.

The displayed P80 date means 80% of the simulated outcomes completed on or before that date under the configured assumptions.

Comparable-ticket score

The score is used only to choose relevant historical tickets; it is not a developer performance score. A historical ticket receives up to 17 similarity points: same issue type (4), same priority (1), shared components (up to 4), shared labels (up to 3), and overlapping summary terms (up to 5). The engine samples real durations from the 20 closest tickets.

Near-term roadmap

  • Direct Google Calendar and Microsoft 365 OAuth connections. Provider-neutral ICS synchronization is available in the MVP.
  • Jira OAuth 2.0 with PKCE or a small authorization service.
  • Dependency and blocker simulation from Jira issue links.
  • Per-team model calibration and forecast backtesting.
  • Optional local text embeddings for stronger comparable-ticket retrieval.
  • Markdown forecast snapshots and Dataview-friendly properties.