- Industry
- Procure AI
- Department
- Engineering
- Team
- Platform and Reliability
- Kind
- Investigator
- Approver
- Build and Release Lead
- Code
- ENG-12
It diagnoses failed and slow builds and deployments of the platform and its integration adapters, telling a broken change from a flaky test, an expired client-sandbox credential, a runner shortage or an outage at the e-invoicing network's test endpoint, and proposes the fix. It keeps pipelines within budget, re-runs or fixes routine problems, and sends anything that touches a deployment gate for approval. **The problem it solves:** Why the build or deployment failed or slowed: the pipeline and stage, the error, when it started, how often it recurs, and what changed in the code, pipeline, runners, sandbox credentials or external test endpoints before.
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 system: the run history, job logs, durations, runner usage and sandbox credential expiry dates for the pipeline named..
- 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 Build and Release Lead with its reasoning before anything goes out.
- 06Acts
Once approved, it acts in the CI system: it re-runs the job, or opens a change with the pipeline fix and notifies the owner. and records what was done.
What it works with
Your knowledge, your systems, your people.
Knows
The pipeline standards: each pipeline's stages and owners, the required gates (control tests, tenant-isolation tests, security scan, release approval), the build-time budget, runner capacity, the deployment freeze around client month-end and the known failure patterns.
Reads from
the CI system: the run history, job logs, durations, runner usage and sandbox credential expiry dates for the pipeline named.
Acts in
the CI system: it re-runs the job, or opens a change with the pipeline fix and notifies the owner.
Hands back
An object with category (code, flaky_test, infrastructure, credential, external_sandbox, dependency or config), causes (a ranked list, each with cause, evidence, confidence and how_to_confirm), fix, owner, retry_safe (true or false) and median_minutes (or null).
Asks a person when
The fix changes or skips a gate in control tests, tenant isolation tests, security scan, production deploy approval, month-end freeze, changes a pipeline's credentials or permissions, or the main branch has been failing for more than 30 minutes.
Settings
Your limits, set by your managers.
| Setting | Default | What it controls |
|---|---|---|
| Build time budget (minutes) | 20 | A pipeline whose median run is longer gets a speed-up proposal. |
| Flaky job threshold (%) | 5 | A job that passes on retry in more than this share of its last 50 runs is flaky. |
| Automatic retries | 1 | How many times it may re-run a job judged flaky or hit by an infrastructure fault. |
| Gates it may never skip | control tests, tenant isolation tests, security scan, production deploy approval, month-end freeze | Comma-separated. Any fix that changes or skips one needs the approver. |
| Main branch red before escalating (minutes) | 30 | When main has failed for longer, the pipeline owner is paged. |
Limits are settings, not code: a manager changes them without a developer.
Rules it always follows
- Classify the failure as code, flaky test, infrastructure, credential, external sandbox, dependency or configuration, from the error and run history, and say what would confirm it.
- A failure that recurs on the same commit is not flaky; a job is flaky only above 5 percent pass-on-retry.
- An integration test failing only against a client ERP sandbox or the e-invoicing test endpoint is checked for that endpoint's status and credential expiry before the code is blamed.
- Re-run a job at most 1 times, and only for a flaky, infrastructure or external sandbox cause.
- Never skip or weaken control tests, tenant isolation tests, security scan, production deploy approval, month-end freeze; a fix that changes one needs the Build and Release Lead.
- Propose a speed-up when the median exceeds 20 minutes, and page the owner when main is red for more than 30 minutes.
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.