Automation pipeline · Launching soon
Evidence follows every stage.
Explore Operations Center’s seven-stage Ansible pipeline: reuse, blueprint, quality, runtime QA, hardening, compliance and pull-request delivery.
A request enters. A pull request leaves.
The pipeline turns a scoped IT request into a candidate change that a person can inspect. Stage numbering follows the product’s implementation, from 0 to 6.
0 · Reuse detection
Existing roles, playbooks and collections are searched and scored before anything is generated. Reuse is preferred over generation.
- Six-factor similarity scoring across the automation catalog
- Exact matches returned immediately, near matches extended rather than rebuilt
- Match confidence reported to the requester
1 · Blueprint builder
A repository and role design is produced before code: structure, variables, tasks, handlers, tests, job template and a documentation plan.
- Repository structure
- Role and variable design
- Handlers and task breakdown
- Molecule scenarios and documentation plan
- AWX job template
2 · Quality gate
Architecture review plus static analysis in one pass: yamllint and ansible-lint, naming, modularity and idempotency expectations.
- ansible-lint and YAML validation
- Architecture review against written standards
- Naming and modularity checks
- Idempotency expectations
3 · Runtime + AWX QA
Molecule scenarios in disposable containers, then against real managed nodes over SSH and WinRM, then a live AWX job launch. A failure in any phase feeds an automated fix loop.
- Molecule scenarios in Docker
- Molecule against real managed nodes (SSH and WinRM)
- Live AWX job execution, stdout captured
- Test logs and job artifacts retained
4 · Hardening
Security review before a human ever sees the change. This stage can block delivery.
- Secret detection
- CVE and dependency review
- CIS and risky-module checks
- Secure defaults; secret-manager lookups where configured
5 · Compliance
Governance gates and evidence retention. Runs in parallel with hardening, and can also block delivery.
- Data anonymization checks
- AI governance record and risk classification
- Privacy and auditability review
- Evidence retained against the change
6 · Deploy — pull request delivery
A Git-based pull request carrying the full evidence trail, opened for human review, with the work item updated and artifacts linked. This is the pipeline's result: nothing reaches production without peer review and merge.
- Pull request opened in your Git platform
- Full evidence attached to the change
- Work item transitioned to peer review
- No automatic production deployment
Review the evidence
Quality, hardening and compliance reports accompany runtime evidence, including AWX job output. The interface must distinguish a proposed change from an approved and deployed one.
Let’s talk