This packet belongs to the alert-to-runbook-revision records guide, which explains how the links work. This page is the form itself. Once it is filled, the completeness checklist helps you judge whether the answers describe a trail worth discussing.
The packet rule: describe the container, not the contents
Every field answers one of five questions: what, how much, how long, which system, who owns it. If an answer would require someone to open an item and read it, the field does not belong on the form.
| Allowed on the form | Kept off the form |
|---|---|
| "Monitoring tool in use since 2020" | Check names, queries or thresholds |
| "Thousands of paging events a year" | Event titles, notes or responder names |
| "Write-ups carry a reference number since 2021" | Any write-up text or stated cause |
| "Procedure edits usually have a comment" | Procedure steps, commands or host names |
| "Notes may mention customers: unknown" | Which customers, or any customer data |
The right-hand column is not a list of things to handle carefully. It lists things a partner never requests. The Confluence inventory template applies the same line to a single wiki.
Packet fields, block by block
Fill in one block per tool, then one block for the links between them. The third column says where the sponsor usually finds each answer without opening anything.
| Block | Field | Where the answer comes from | Usually confirmed by |
|---|---|---|---|
| Monitoring | Tool and years in use, earlier tools if any | Admin settings, contract dates | Reliability lead |
| Monitoring | Object types retained: checks, trigger and recovery events | Product configuration | Reliability lead |
| Paging | Tool and years in use; event history retained? | Account settings, retention policy | On-call manager |
| Paging | Volume band per year | Account reporting dashboards | On-call manager |
| Wiki | Space for procedures and write-ups; first year | Space settings, oldest page date | Platform manager |
| Wiki | Volume band of procedures and write-ups | Space page counts | Platform manager |
| Links | Monitoring check to paging event | Integration configuration | Reliability lead |
| Links | Paging event to write-up | Write-up template fields | Platform manager |
| Links | Write-up to procedure edit | Team habit on edit comments | Platform manager |
| Links | Does a recovery close the paging event automatically? | Notification templates | Reliability lead |
| Gaps | What is unknown, and who could find out | The sponsor's own judgment | Named per item |
Two fields carry most of the value. The years tell SourceX how deep the history runs; the joins tell it whether those years form one connected story or three separate piles. The PagerDuty incident export page covers what that tool's own records contain.
A filled packet (Illustrative)
A fictional 300-person payments software company, answered by its SRE lead and platform engineering manager in one call. Every value is Illustrative.
| Packet field | Entry | Owner who can confirm |
|---|---|---|
| Monitoring | Datadog since 2020; an earlier tool for 2017 to 2019 | SRE lead |
| Paging | PagerDuty since 2018; incident log entries retained | On-call manager |
| Knowledge | Confluence runbooks and postmortems since 2017 | Platform engineering manager |
| Volume band | Thousands of incidents; a few hundred postmortems; about 150 runbook pages | SRE lead |
| Monitoring-to-paging link | Paging handle on most production monitors | SRE lead |
| Paging-to-write-up link | Incident number field in the postmortem template since 2021 | Platform engineering manager |
| Write-up-to-procedure link | Version comments on runbook edits: inconsistent | Unknown until reviewed |
| Recovery behavior | Recovery messages include the paging handle | SRE lead |
| Monitoring migration | No mapping kept between old and new monitors | SRE lead |
| Unknown | Whether incident notes include customer data | General counsel |
Read the packet the way a reviewer would. Alerts and incidents are joined by design from 2020. Incidents and postmortems are joined by a template field only from 2021. The procedure link depends on habit and is uneven. The 2017 to 2019 alert history exists but sits in a tool with no mapping to the current one. That is an honest, specific description of nearly a decade of operating history, and it took no record access to write.
Fields that change how the packet is read
Three entries look minor and are not.
Recovery behavior. Datadog's PagerDuty integration resolves the linked incident automatically when a monitor recovers, if the paging handle is in the recovery message. Datadog builds that message with conditional variables such as {{#is_recovery}}, so the sponsor can answer from the message templates without opening an alert. If recoveries auto-resolve, some incidents will have closed while the fault was still present, and the real cause sits only in postmortems. The missing context page shows what that looks like in a single export.
Incident log retention. PagerDuty keeps a log of what happened to each incident: integration event data, notifications, who acknowledged or resolved it, automatic actions, reassignments and notes. Whether that log is still held for older years decides whether response steps can be reconstructed, so ask for the earliest year it covers.
Version comments. Confluence page history lists each version's date, author and comment. A comment that cites a postmortem or an incident number is the only thing that ties a runbook change to its cause. "Usually", "sometimes" and "rarely" are all useful answers; so is "since the template changed in 2022". The checklist for document revision histories covers version records in more depth.
Write each answer at the confidence the sponsor gave it. "Unknown until reviewed" is a complete entry, provided it names who could find out.
Who answers the packet, and what it does not decide
The SRE lead, on-call manager and platform engineering manager answer the fields. None of them can authorize a licensing discussion: that is the CEO, usually with the CTO and general counsel involved, because incident notes and runbooks can mention customers, vendors and security controls. Context on why operational records like these matter to AI developers is in the AI agents in ITSM guide.
Before the packet goes anywhere, confirm the company baseline: 50+ full-time employees at peak (contractors excluded), documented operating history, rights to license, 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.
The packet does not decide whether the company's rights cover the records, what must be removed for security, whether a buyer wants this kind of history, or what the company would be paid. SourceX settles those under the company's own agreement. SourceX licenses approved business data directly from companies for AI training and evaluation, and manages agreement, rights review, data preparation, delivery and payout. It does not claim a connector for Datadog, PagerDuty or Confluence, and a partner should not promise one. Sign the Channel Partner Agreement before you submit an introduction; attribution goes to the first valid introduction within the 90-day window.
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 the SRE lead can fill in this packet and an officer will take the first call, register as a partner, check the company against who qualifies, and read the terms before you send it.