Configure Integrations between sfdx-hardis and external tools
sfdx-hardis works out of the box without any integration: everything it reports is available in the job logs and in the terminal.
Configuring integrations with external tools makes that information easier to reach: deployment results posted directly in Pull Requests, notifications sent to Slack, Microsoft Teams, Google Chat or Email, links to your tickets, and AI assistance to solve deployment errors.
Git Providers
Depending on your git provider, configure one of the following integrations.
- GitHub
- Deployment status in Pull Request comments
- Quick Deploy to improve performance
- GitLab
- Deployment status in Merge Request notes
- Quick Deploy to improve performance
- Azure Pipelines
- Deployment status in Pull Request threads
- Quick Deploy to improve performance
- Bitbucket
- Deployment status in Pull Request comments
- Quick Deploy to improve performance
Pull Request comments
Whatever the git provider, sfdx-hardis can post up to three different comments on the same Pull Request. Each one starts with a banner telling which comment you are reading, and how it went: green when it succeeded, red when it failed, orange when manual actions are still waiting. The banner replaces the title of the comment, which stays as the alt text of the image.
Validation
Posted by the validation job, before the merge. It contains the result of the deployment simulation on the target org: deployment errors, Apex test results, code coverage and Flow visual diffs.
Deployment
Posted by the deployment job, after the merge. It contains the result of the real deployment in the target major org.
Deployment actions
Updated by both jobs. It tracks the deployment actions of the Pull Request: which action ran in which org, and the checkboxes of the manual actions that are still waiting to be performed.
Navigation between the comments
The three comments are posted by different jobs and end up scattered among the other comments of the Pull Request, so each one displays a navigation line above its banner:
🔍 Validation | 🚀 Deployment | 🛠️ Actions
The comment you are reading is in bold, the others are links to their own comment. Comments that do not exist yet are not listed. The same navigation is added at the very beginning of the Pull Request description, so the comments can be reached from the top of the Pull Request.
The description is only modified between the sfdx-hardis navigation markers: the text written by the author is left untouched.
On Azure DevOps, the description of a completed Pull Request can no longer be edited, so the validation job creates the deployment comment right away, with a "Waiting for the Pull Request to be merged" note and no banner. The deployment job updates that same comment once the merge happened, which keeps the Deployment entry of the navigation valid from the start.
Environment variables
| Variable | Effect when set to false |
|---|---|
SFDX_HARDIS_PR_COMMENT_BANNERS |
Post the comments without their banner image, for example when the network blocks raw.githubusercontent.com |
SFDX_HARDIS_PR_COMMENT_NAV |
Post the comments without the navigation line, and never modify the Pull Request description |
SFDX_HARDIS_PR_DESCRIPTION_NAV |
Keep the navigation in the comments, but never modify the Pull Request description |
Message notifications
- Slack
- Notifications
- Microsoft Teams
- Notifications
- Google Chat
- Notifications
- Email
- Notifications
- API (e.g. Grafana)
- Notifications
Ticketing providers
- Jira
- Enrich Pull Request comments with ticket references and links
- Enrich notifications with ticket references and links
- Post a comment and a label on Jira issues when they are deployed in a major org
- Azure Boards
- Enrich Pull Request comments with work item references and links
- Enrich notifications with work item references and links
- Post a comment and a tag on Azure work items when they are deployed in a major org
- Generic ticketing
- Enrich Pull Request comments with ticket references and links
- Enrich notifications with ticket references and links
Large Language Models (AI)
- Agentforce
- Deployment Agent
- Project Documentation
- OpenAI, Anthropic, Gemini, Ollama
- Deployment Agent
- Project Documentation



