## **Is your feature request related to a problem? Please describe.** Our GitLab instance uses a certificate from a custom internal CA. Adding an option to add this CA as trusted would be very nice. <!-- Describe the problem this feature would solve. Be specific about the attack vector, compliance gap, or UX issue. Include links to relevant documentation, CVEs, or real-world examples when possible. Example: "When CI_DEBUG_TRACE is enabled, all environment variables including masked secrets are printed in job logs. This is a well-documented attack vector..." --> ## **Describe the solution you'd like** Add a `PLUMBER_TRUSTED_CA` variable or config which we can point to the file with the chain. <!-- Describe the feature you'd like to see. For new controls, include: - The control name (e.g., pipelineMustNotDoX) - What it detects - Example YAML snippets showing what would be flagged For other features, describe the expected behavior and user-facing interface (CLI flags, config options, output format, etc.) --> ### Configuration in `.plumber.yaml` ```yaml controls: controlName: enabled: true # Add relevant configuration options here ``` ## **Implementation Hints** <!-- Optional. Keep it at the level of WHAT the rule should detect and WHICH data it needs. Don't prescribe specific Go files — see CONTRIBUTING.md ("Adding a New Control") for the full file-by-file implementer checklist that walks every layer (collector + IR, Rego rule, config plumbing, catalog, bench gate, terminal output, JSON output, PBOM/CycloneDX, compliance, --skip-controls, plumber config validate, tests, docs). That checklist is copy-pasteable into Claude Code / Cursor / Copilot so the implementer (or their LLM) can tick boxes step by step. Detection logic for every new control lives in a Rego rule under `policies/*.rego`. The `control/` package is bookkeeping only; the pre-Rego Go-function pattern has been fully phased out. Consider covering (briefly): 1. Data source: which IR field would the rule read? If the data isn't on the IR yet, does it need collector enrichment (GitLab/GitHub API call, repo-artifact scan, …)? Note: Plumber executes no external binaries during analysis (#310). 2. Logic: a one-liner of what the rule flags. 3. Severity + ISSUE code suggestion (or "next available"). 4. Default state: opt-in or default-on? --> ## **Why It's Valuable** <!-- Explain the impact. Why should this be prioritized? Who benefits? Link to OWASP CI/CD risks, real incidents, or competitor features when relevant. --> > **Note:** If you submit a PR for this feature, please keep "Allow edits from maintainers" enabled so we can collaborate more easily.
Activity