Audit CI pipeline evidence integrity

What it's for

Turns Gitrevio's CI failure-rate card into a trustworthy assurance ledger: which required changes really ran, which jobs and reruns belong to them, whether evidence is complete, and where feedback capacity is leaking.

What you give it

Inputs split into evidence read from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.

{
  "as_of_ms": 8640000000,
  "eligible_changes": [
    {
      "commit_sha": "sha-a",
      "evidence_verified": true,
      "id": "change-a",
      "merged_at_ms": 8208000000,
      "repository_id": "repo-a",
      "requires_ci": true
    }
  ],
  "jobs": [
    {
      "allowed_failure": false,
      "attempt": 1,
      "evidence_verified": true,
      "finished_at_ms": 8208062000,
      "id": "job-a",
      "job_name": "tests",
      "pipeline_run_id": "run-a",
      "required": true,
      "started_at_ms": 8208002000,
      "status": "success"
    }
  ],
  "pipeline_runs": [
    {
      "commit_sha": "sha-a",
      "configuration_hash": "cfg-a",
      "evidence_verified": true,
      "finished_at_ms": 8208062000,
      "id": "run-a",
      "pipeline_name": "main",
      "repository_id": "repo-a",
      "rerun_of_id": null,
      "source_complete": true,
      "started_at_ms": 8208002000,
      "status": "success",
      "triggered_at_ms": 8208001000
    }
  ]
}

What you get back

This is the actual output of running the example above — computed by the same function the platform calls, not an illustration.

{
  "decision": "accepted",
  "detail_truncated": false,
  "finding": "evidence_ready",
  "gates": {
    "change_coverage_passed": true,
    "flake_proxy_rate_passed": true,
    "integrity_passed": true,
    "job_coverage_passed": true,
    "maximum_flake_proxy_rate": 0.1,
    "maximum_unlinked_job_rate": 0.01,
    "minimum_change_coverage": 0.95,
    "minimum_job_coverage": 0.98,
    "unlinked_job_rate_passed": true
  },
  "integrity_diagnostics": {
    "ambiguous_change_keys": 0,
    "ambiguous_run_links": 0,
    "attempted_job_groups": 1,
    "duplicate_job_attempts": 0,
    "flake_proxy_groups": 0,
    "future_changes": 0,
    "future_job_information": 0,
    "future_run_information": 0,
    "future_runs": 0,
    "impossible_job_times": 0,
    "impossible_run_times": 0,
    "incomplete_run_sources": 0,
    "inconsistent_success_runs": 0,
    "missing_job_runs": 0,
    "out_of_window_jobs": 0,
    "prewindow_changes": 0,
    "prewindow_runs": 0,
    "rerun_cycles": 0,
    "rerun_mismatches": 0,
    "unfinished_terminal_jobs": 0,
    "unfinished_terminal_runs": 0,
    "unknown_rerun_parents": 0,
    "unlinked_job_rate": 0,
    "unlinked_jobs": 0,
    "unlinked_runs": 0,
    "unverified_changes": 0,
    "unverified_jobs": 0,
    "unverified_runs": 0

How it works

Statistical audit & measurement — Check whether a number is fit to decide on: coverage, timing, reconciliation, and the gaps a dashboard hides.

1. 1 Freeze the complete CI-required change cohort and as-of cutoff, then reconcile repository and commit identities to every in-window pipeline run and job while preserving future, pre-window, ambiguous and orphan records as diagnostics.
2. 2 Validate run/job timestamps, terminal completeness, required-job consistency, source verification, rerun parent acyclicity and exact commit/configuration/pipeline continuity; calculate change and job coverage before any performance metric.
3. 3 Measure aggregate feedback, compute and same-change same-configuration fail-then-pass flake proxies only after lineage passes; separate evidence acceptance from a high-flake review finding and never infer individual fault or authorize a branch-policy change.

Before you trust it

Every tool in the catalog ships with the conditions under which its answer is meaningful — and the conditions under which it should abstain instead of guessing.

Calibrating it to your org

Calibration workflow

1. 1 Define the management decision, target outcome, aggregate unit, privacy boundary, cadence, and prediction/intervention horizon for this organization.
2. 2 Build a tenant-scoped historical cohort using only information available before each prediction or decision; preserve zero periods, censoring, assignment probabilities, and unresolved outcomes when the method requires them.
3. 3 Estimate statistical parameters on training history, but obtain costs, utilities, risk tolerance, practical-effect thresholds, capacity, and policy constraints from accountable decision owners.
4. 4 Validate on later time windows or held-out aggregate units at the deployment grain, against a simple baseline and the function-specific validation strategy.
5. 5 Deploy only if the returned decision clears evidence, overlap, calibration, robustness, and guardrail checks; warning, unsupported, schema-gap, and fallback decisions are abstentions.
6. 6 Monitor realized outcomes, data drift, coverage, and decision regret; recalibrate at a governed cadence or after a detected regime/definition change, never merely because a stakeholder dislikes the result.

Call it from your AI

You don't wire up 388 tools in your MCP client. The GitRevio MCP server exposes 18 tools, three of which let an agent search the catalog, read a tool's schema, and run it — so the assistant finds this one on its own.

gitrevio_capabilities_search
  { "q": "audit the complete pointintime changetopipelinetojobtorerun cohort" }
  → finds "audit_ci_pipeline_evidence_integrity"

gitrevio_capability_describe
  { "capability_id": "audit_ci_pipeline_evidence_integrity" }
  → returns the input schema and agent guidance shown on this page

gitrevio_capability_run
  { "capability_id": "audit_ci_pipeline_evidence_integrity", "arguments": { ... } }
  → returns the result shown above

Works in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.

Related tools

See every tool in Quality, incidents & reliability →