/
/
Know when a BugHerd fix is live before you retest
For project and QA leads who want to know if a reported bug's GitHub fix has been committed and deployed, just by asking their AI assistant.
On this page
Connector
Connect BugHerd, pick your AI tool and paste the prompt.
The problem
Nobody knows if a reported fix is actually live
To tell a client whether a bug is fixed, someone finds the linked GitHub issue, searches commits and checks whether a workflow run deployed them. Retest too early and you waste a round trip, and an undeployed fix looks like no fix at all.
Without PopdockAI
- Find the GitHub issue linked to the BugHerd task
- Search commits for the issue or task number
- Check workflow runs for a deploy that included them
- Retest too early and send the client back and forth
With PopdockAI
- The task, archived or not, is traced to its GitHub issue
- Fixing commits are found by issue or task number
- Deploy runs on your branch are checked for success
- A clear verdict with evidence, read-only
The prompt
Use this prompt
Copy it into any AI tool connected to PopdockAI. Swap the values in braces for your own.
**Objective:** Tell whether the fix for task `{task-number}` in the BugHerd project `{project-name}` has been committed to `{repository}` and deployed from the branch `{branch}` (default main), so the task can be retested.
How the two systems are linked: a BugHerd task created from a GitHub issue carries "<repository>#<issue number>" in its external-id. A GitHub issue created from BugHerd starts its title with "[BugHerd <project name> #<task number>]".
**Step 1: Read the task**
Call `bugherd-get-projects` with `name.contains` set to `{project-name}`. If nothing matches, say so and stop. If several match, list their names and project-id values and ask which one; do not guess. Then call `bugherd-get-tasks` with `project-id` and `task-number` (then `bugherd-get-archived-tasks` if not found). If it is not found, stop.
**Step 2: Find the GitHub issue**
If the external-id is "{repository}#<n>", call `github-get-issue` with `company` and `issue-number` n. Otherwise call `github-get-issue` with `company` and `title.contains` set to "[BugHerd" plus the project name and " #" plus the task number. Note the issue's status and closed-date. If there is no issue, carry on with Step 3 using the task number alone.
**Step 3: Find the fixing commits**
Call `github-get-commit` with `company` and `message.contains` set to "#<issue number>" when there is an issue, and again with `message.contains` set to "BugHerd #<task number>". Keep commits dated after the task was created.
**Step 4: Check the deploy**
Call `github-get-workflow-run` with `company`, `branch` set to the branch and `created-date.from` set to the earliest fixing commit's commit-date. A run whose conclusion is success and whose sha is a fixing commit, or that started after the last fixing commit, has shipped the fix.
**Step 5: Report**
* **Verdict:** shipped, committed but not deployed, deploy failed, or no fix found.
* **Evidence:** issue (number, status, issue-url), commits (sha, author, date, first line of message), workflow run (name, run-number, conclusion, date, url).
* **Next step:** retest the page on the site, wait for a deploy, or chase the developer.
**Rules & Guidelines:**
* Tools named bugherd-... are on the bugherd MCP server; tools named github-... are on the github MCP server. Every GitHub tool takes the repository as its `company` parameter. If either set of tools is missing, say which server is not connected and stop.
* This prompt is read-only. Do not call any BugHerd write tool (`bugherd-add-task`, `bugherd-add-feedback`, `bugherd-update-task`, `bugherd-move-task`, `bugherd-add-comment`, `bugherd-add-attachment-from-url`, `bugherd-add-project`, `bugherd-update-project`, `bugherd-add-project-member`, `bugherd-add-guest-to-project`) or any GitHub write tool (`github-add-issue`, `github-update-issue`, `github-add-issue-comment`, `github-update-issue-comment`, `github-add-issue-assignee`, `github-add-issue-label`, `github-lock-issue`, `github-unlock-issue`, `github-add-milestone`, `github-update-milestone`, `github-enable-workflow`, `github-disable-workflow`, `github-rerun-workflow-run`, `github-cancel-workflow-run`).
* A commit mentioning the number is evidence, not proof; say so when the message is ambiguous.
**Validation:**
* If no issue and no commit mention the task, the verdict is "no fix found".Who it's for
One answer, three teams that use it
Project managers
Tell clients when to retest
Answer is it fixed yet with evidence, so clients only retest once the fix is deployed.
Is BugHerd task 88 in the Contoso Portal project fixed and deployed? Repo contoso/portal, branch main.
QA testers
Retest only what has shipped
Check that a fix reached the site before spending time on a retest.
Has the fix for task 41 in the Fabrikam Shop BugHerd project been deployed from fabrikam/shop main?
Developers
Find fixes stuck before deploy
Spot fixes that are committed but not deployed, or whose deploy failed, so they do not sit unnoticed.
Check whether BugHerd task 305 in the Northwind Traders site project is committed but not deployed in northwind/web.
How it works
What PopdockAI does when you ask
You give the task and the repository, and your AI assistant traces the fix through GitHub. It only reads, and says when a commit message is loose evidence.
01
Ask
You give the BugHerd project, task number and GitHub repository.
02
Trace
It reads the task and finds the linked GitHub issue by its reference or title.
03
Check
It finds commits that mention the issue or task, then looks for a successful deploy run on your branch.
04
Report
It gives a verdict with the issue, commits and run as evidence, and the next step.
What PopdockAI provides
The tools behind the answer
Ready made tools for BugHerd and one secure connection to your AI tool. No code, no exports.
bugherd-get-projects
bugherd-get-tasks
bugherd-get-archived-tasks
github-get-issue
github-get-commit
github-get-workflow-run
This report is built from Get data tools only. PopdockAI applies the row, field and parameter rules your admin sets, and every query is written to the audit trail.
The outcome
Know a fix is live before you retest
Related use cases
More ways to put this data to work
One prompt template with infinite possibilities.
Your AI assistant adapts its approach, tone, and capabilities to match each unique context while you maintain a single, manageable workflow.
Join forward-thinking companies that are transforming their operations with intelligent, secure AI automation.

.webp)
