# Incident-to-fix missing context: before and after | SourceX

Canonical: https://refer.sourcex.si/guides/incident-to-fix-missing-context
Last updated: 2026-11-04

A single-system incident export usually keeps the conversation and loses the decision. A Slack export alone shows when the incident started, who responded and when someone said it was resolved. It does not show the severity the team assigned, which code changed, whether anyone reviewed the change, or whether the first fix actually worked. That context lives in the issue tracker and the code host, and it only reconnects if the same identifier was typed in all three places.

This page is part of the [incident-to-fix guide](/guides/chain-incident-to-fix). It reads one Illustrative incident twice, first as a Slack export tells it and then with Jira and GitHub beside it, so a partner can see what to ask about before describing an incident history to anyone.

## Before: the incident as a Slack export tells it

Illustrative: a 260-person logistics software company. Every name, time and number is fictional.

The export holds one folder for a private channel called inc-search-down, with a JSON file for the day. Read message by message, it says:

- **09:14.** An engineer opens the channel: shipment search returns empty results for some customers.
- **09:20.** Two more engineers join. One posts a dashboard link.
- **09:41.** Someone writes "rolling back the index mapping change."
- **10:05.** "Rollback deployed, results look normal."
- **10:30.** The incident lead posts: "Resolved. Thanks all."
- **11:52.** One more message: "seeing a few stragglers, on it." No reply follows in the channel.

A reader of this export alone would conclude that a rollback fixed the problem in about an hour and that a few leftover errors were cleaned up informally. The dashboard is a link, not a file. The severity is never stated. The rollback has no change number, and nobody in the channel says who checked it.

## After: the same incident with Jira and GitHub beside it

Put the same morning next to the issue tracker and the code host, and three things change.

| Time | System | What the connected history adds |
|---|---|---|
| 09:22 | Jira | INC-733 opened at severity 2, key posted in the channel |
| 09:48 | GitHub | PR 2210 on branch INC-733-revert-mapping, one approval recorded |
| 10:02 | GitHub | PR 2210 merged to main |
| 11:58 | GitHub | PR 2231 on branch fix-reindex, merged to a release branch with no approval |
| 13:15 | Jira | INC-733 resolved, linked to PR 2210 only |
| Next day | Shared drive | Postmortem names a stale reindex job as the root cause and cites INC-733 |

Now the story is different. The rollback stopped most of the errors, but the root cause was a reindex job, fixed by a second change merged 88 minutes after the resolved message. That second change carries no issue key, so Jira never linked it, and it went to a release branch where nobody approved it. "Seeing a few stragglers" was the actual fix.

## What goes missing, and where that context lives

| Lost context | Where it lives | How it rejoins the chain | Survives a standard export? |
|---|---|---|---|
| Severity and priority | Jira issue fields | Issue key posted in the channel | Depends on the company's Jira export, ask the administrator |
| Which code changed | GitHub pull requests and commits | Issue key in branch name, commit or PR title | Depends on how the company archives its repositories |
| Whether the change was reviewed | GitHub approvals on the pull request | Pull request number | Only where review rules or habits produced approvals |
| Whether the first fix worked | A later merge in the incident window | Issue key, if typed; otherwise time only | The merge survives; the link may not |
| Root cause | Postmortem document | Issue key, if the author cited it | Depends on where postmortems are stored |
| Dashboards and screenshots | Monitoring tools, file storage | A link in the thread | No; Slack exports carry files as links |

The rejoin column is the important one. Atlassian's documentation says the work item key in a branch name, commit message or pull request title links development work to the Jira item in connected tools such as GitHub. Where engineers followed that convention, the after view assembles itself. Where they did not, as with PR 2231 above, the context exists but floats free. The [record chain page](/guides/incident-to-fix-record-linkage) lists every key in the sequence step by step.

Review evidence is a configuration question, not a given. GitHub's protected-branch settings can require a set number of approving reviews before merging and can dismiss stale approvals after new commits, but only on the branches where an administrator applied them. A reviewed patch on main and an unreviewed hotfix on a release branch can sit inside the same incident.

## How much of the before view a company can even export

Before a partner talks about coverage, two facts matter more than how old the workspace is: the Slack plan and the channel type.

Slack's help center says exports hold one folder per conversation with daily JSON files, each message carrying the sender's user ID and a timestamp, and files appear as links. On Free and Pro plans, owners and admins can export public channels; private channels and DMs are exportable only in limited circumstances after an application. Business+ and Enterprise owners can apply for a broader self-serve export. A company that ran every incident in private channels on a Pro plan may have little of the before view to offer, even with years of incidents.

Slack's API terms are a separate question from export rights, covered in [the Slack API terms answer](/questions/slack-api-terms-llm-training-restriction). Teams that page through an incident tool first also hold alert and acknowledgment times there; the [PagerDuty export page](/systems/pagerduty-export-incidents) covers that system.

## Describing the gap in an introduction

The partner's job is to describe which systems hold which part of the story, not to rebuild it. Useful notes: where incidents are handled, which plan the workspace is on, whether the issue key is posted in incident channels as a habit, whether release branches have review rules, and where postmortems live. Leave every message, issue description and diff where it is.

The DevOps or platform lead can answer those questions. The CTO or another executive with authority over company data decides whether to discuss a license, and counsel reviews customer contracts, because incident threads often quote customer details. The company needs 50+ full-time employees at peak (contractors excluded), a documented operating history, rights to license the records and an authorized contact. Companies in the U.S., Canada and the EU qualify, as do companies in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia when at least 70% of the data is in English.

Whether the joined history survives export, whether customer data limits its use, and whether anyone wants it all stay unknown until SourceX reviews it with the company. For baseline screening of incident records, the [incident and on-call introduction checklist](/resources/a-company-introduction-checklist-for-incident-and-on-call-records) is the starting point; the [completeness scorecard](/resources/incident-to-fix-completeness-checklist) grades the chain in more detail.

## Next step

> **Refer a company. No cap on rewards.** Default reward on a company's first qualifying transaction: the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap; further qualifying transactions first agreed within 12 months of the first Closing pay 15%. Eligibility and payment conditions apply, including the USD 100,000 Minimum Transaction, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.

If a client can show that its incident threads, issues and merges share one key, so the after view is recoverable, [register as a partner](/register), check the company against [who qualifies](/who-qualifies), and read the [program terms](/terms) before you introduce its CTO.

## Sources

- https://slack.com/help/articles/220556107-How-to-read-Slack-data-exports
- https://slack.com/help/articles/201658943-Export-your-workspace-data
- https://support.atlassian.com/jira-software-cloud/docs/reference-issues-in-your-development-work/
- https://docs.github.com/repositories/configuring-branches-and-merges-in-your-repository/defining-the-mergeability-of-pull-requests/about-protected-branches
- https://refer.sourcex.si/terms
