Outsourced ProgrammersPowered by StealthAgents

How many outsourced mobile builds share one signing key?

A protocol for counting signing keys, mapping each key to its builds, and reporting the ones with no rotation record.

How many outsourced mobile builds share one signing key?
WORK_LANE / RESEARCH BRIEF / MOBILEREVIEW QUEUEACCEPTANCEPASSowner review requiredEVIDENCE TRAILTICKET → BRANCH → REVIEW → CLIENT RELEASE
Table of contents
  1. Research context
  2. Practical application
  3. Operating controls
  4. Sources
  5. Frequently asked questions

Research context

This protocol measures how many mobile builds depend on one signing key in an outsourced delivery lane. The unit is one key on one platform account at the observation date. A shared key means two or more apps or build tracks use the same key identifier. The study describes key concentration and custody, not the security of the apps themselves.

Define the platform accounts and the apps in scope before collection. Include the store accounts, the internal distribution accounts, and any enterprise account the team can reach. Record the account owner and the recovery contact for each. Publish the list with the observation date so a later reviewer can reproduce it.

Define a key by its identifier, its certificate fingerprint, and its platform. Record the apps and tracks that use the key, and the build identities allowed to sign with it. Note the storage location and who can read the key material. Test the definition on the first two accounts.

Observe the store console and the build pipeline rather than the key inventory alone. Look for a build identity that can sign with a key that is not in the inventory. Record the discovery source for each mismatch. The study should report the keys that the inventory missed.

Count the builds and the apps per key for the observation window. A key used by one app on one track is different from a key used by several apps across several tracks. Record the count and the platform. Report the keys with no recorded rotation and the keys with no named custodian.

Practical application

Check whether the client or the contractor controls the keystore and the recovery material. Record the account, the storage system, and the person who can export the key. A key held in a personal account is reported separately from one held in a client secret store. The distinction matters more than the raw count.

Report the number of keys, the number of apps, and the number of builds that depend on each key. Show the counts by platform and by account type. Present the keys with no rotation record and the keys with no custodian separately. Use the same denominator for every rate.

Separate a shared key from a shared account. A shared key signs several apps, while a shared account lets several people use one identity. Report the two categories separately because the fixes differ. Record the evidence beside each finding.

Inspect the keys around a team change or a store policy change. A cluster of unrotated keys may mean the rotation step depended on the engineer who left. Compare the pattern with the personnel record. A single stale key can be an oversight, while a cluster is a continuity gap.

Do not treat a key that still works as proof that custody is sound. The key may be valid while the storage location is a personal laptop or a shared drive. Sample one key and trace who can read it. Report what the trace showed.

Operating controls

The record cannot show whether a key was copied or misused. It can show only whether the key exists, who can reach it, and which builds depend on it. A shared key is not proof of a compromise, and a unique key is not proof of good custody. The protocol also cannot see a key held by a store on the client behalf.

A useful additional view is the number of days since the last rotation for each key. A key that has never been rotated is a different finding from one rotated last quarter. Record the rotation date and the approving person. That field turns a count into a review of lifecycle practice.

Give the result to the client security owner and the offshore delivery lead with one observable change. The change might be a client-held keystore, a per-app key, a rotation calendar, or a build-log check. Record the owner, the effective date, and the measure that will show whether the change worked. Repeat the protocol with the same definitions after the change.

Close the review with one change and a date to re-measure. The change should address the largest shared-key or unowned-key group, not the easiest app to report. Record the expected effect and the evidence that will prove it. A custody note that is never tested is a document, not a control.

Sources

  1. OWASP Mobile Application Security
  2. Android application signing
  3. NIST SP 800-57 Part 1
  4. CISA Secure by Design

Frequently asked questions

What should happen first?

Begin with a bounded ticket, approved access, explicit acceptance evidence, and a named reviewer.

Who approves production changes?

The client's technical owner keeps final merge and release authority.

Build a programming lane your technical owner can review.

Bring the stack, backlog, acceptance checks, and release boundaries. We will help scope the role around the work.

Discuss your programming needs