- Industry
- Banking
- Department
- Engineering
- Team
- Platform and Reliability
- Kind
- Investigator
- Approver
- Head of Technology Operations
- Code
- ENG-12
It diagnoses failed and slow builds and deployments, telling a broken change from a flaky test, an expired credential, a runner shortage or a dependency outage, and proposes the fix. It never bypasses the bank's protected gates: the four-eyes approval for production, the security scan, and the change freeze around month-end and payment peaks. **The problem it solves:** Why the builds or deployments named failed or slowed: which step, since when, what changed in code, tests, credentials, runners or dependencies, and whether it is consistent or intermittent.
How it works
6 steps, and a person signs off before step 6.
- 01Looks it up
Finds the record the request is about in the CI pipeline: the run logs, timings, test history and pipeline changes..
- 02Reads the request
Works out what is being asked and what a complete answer needs.
- 03Checks the facts
Checks figures, dates and rules against the source records, not from memory.
- 04Works it through
Weighs competing explanations against the evidence and shows which one holds up.
- 05Asks a person
Hands the work to Head of Technology Operations with its reasoning before anything goes out.
- 06Acts
Once approved, it acts in the CI pipeline: it applies the approved pipeline fix or retries. and records what was done.
What it works with
Your knowledge, your systems, your people.
Knows
The pipeline standards: protected gates, change freezes (month-end, year-end, payment peak days), the four-eyes production approval, credential rotation, build-time budgets and flaky-test rules.
Reads from
the CI pipeline: the run logs, timings, test history and pipeline changes.
Acts in
the CI pipeline: it applies the approved pipeline fix or retries.
Hands back
An object with runs (a list with run_id, cause, evidence and confidence), fix (a list of steps), flaky_tests (a list), gate_changes (none or a description) and freeze_status.
Asks a person when
The fix changes a protected gate, rotates a production credential, needs an emergency change during a freeze, or main has been red over 30 minutes.
Settings
Your limits, set by your managers.
| Setting | Default | What it controls |
|---|---|---|
| Build time budget (minutes) | 20 | Longer builds are investigated. |
| Flaky threshold (%) | 5 | Tests failing intermittently above this are flaky. |
| Automatic retries | 1 | Infrastructure failures retry this often. |
| Protected gates | four-eyes production approval, security scan, required tests, change-freeze check | Never bypassed. |
| Main branch red alert (minutes) | 30 | Longer is escalated. |
Limits are settings, not code: a manager changes them without a developer.
Rules it always follows
- Never bypass four-eyes production approval, security scan, required tests, change-freeze check.
- Retry infrastructure failures 1 time; a test failing intermittently above 5 percent is flaky.
- A production deployment during a change freeze is refused unless an emergency change is approved.
- Investigate builds over 20 minutes; escalate main red over 30 minutes.
- Rank causes by evidence and say what would confirm each.
See the CI/CD Pipeline Agent agent on your own work.
We’ll run it on a sample of your work and show you what your approver would see.