Add retained, bounded metrics for kani-smoke (#847)

[leynos]

Summary

Record each Kani workflow attempt with fixed decision, outcome, proof duration and cache result labels. An unconditional finalizer writes release-admission JSONL; the workflow uploads it under an attempt-specific artifact name with 30-day retention. Documentation covers bounded vocabularies, cache semantics, operator aggregation, and cancellation and timeout limits.

Closes #847

Validation

• make typecheck-python
• make markdownlint
• make nixie
• make check-fmt
• make test-workflow-contracts — 1,246 passed, 3 skipped; 27 follow-on script tests passed.
• make lint
• make doc-coverage — 98.87% aggregate.
• make test — 3,919 nextest passed, 6 skipped; doctests passed.
• coderabbit review --agent — final review completed with no findings.

Summary by Sourcery

Add retained, bounded observability for every Kani smoke workflow attempt.

New Features:

• Record bounded Kani smoke attempt, cache, outcome, and proof-duration metrics as retained JSONL artifacts.

Bug Fixes:

• Preserve Kani decision and failure observability across scope errors, proof failures, cancellations, and incomplete timing marks.

Enhancements:

• Add fixed decision and outcome vocabularies with documented cache semantics, timing behavior, aggregation guidance, and delivery limits.

CI:

• Wire unconditional Kani metrics finalization and attempt-specific 30-day artifact upload into the CI workflow.

Documentation:

• Document the Kani smoke metrics contract, bounded labels, cache interpretation, timing limitations, artifact retrieval, and aggregation.

Tests:

• Add workflow contract, unit, acceptance, property, duration, and wiring coverage for Kani metrics and retained artifact delivery.

Chores:

• Extend Kani proof scope tracking to include the metrics helper.

References

• Issue #847
• Source PR #774 comment
• Lody session

[coderabbitai]

Important

Draft PR not reviewed

Draft PRs are not automatically reviewed by default.

• Trigger a manual review

To automatically review draft PRs, update your CodeRabbit configuration:

reviews:
  auto_review:
    drafts: true

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

_{Comment @coderabbitai help to get the list of available commands.}

[sourcery-ai]

Reviewer's Guide

This PR adds an attempt-level observability pipeline for kani-smoke: scope decisions, proof and mutation outcomes, cache results, and proof duration are normalized into a bounded JSONL contract, finalized and uploaded even after ordinary failures, with 30-day attempt-specific retention and extensive workflow-contract coverage.

Sequence diagram for Kani smoke metrics finalization

sequenceDiagram
    participant Scope as Scope step
    participant Workflow as kani-smoke workflow
    participant Metrics as kani_smoke_metrics.py
    participant Artifact as Retained artifact

    Scope->>Metrics: write_decision_record()
    Workflow->>Metrics: mark_start()
    Workflow->>Workflow: make kani-ir
    Workflow->>Metrics: mark_end()
    Workflow->>Metrics: finalize()
    Metrics->>Metrics: classify_outcome()
    Metrics->>Metrics: classify_cache()
    Metrics->>Metrics: metrics_records()
    Metrics-->>Workflow: metrics.jsonl
    Workflow->>Artifact: upload metrics.jsonl
    Note over Workflow,Artifact: always(), retention 30 days

Loading

File-Level Changes

Change	Details	Files
Adds bounded, retained metrics emission to the kani-smoke workflow without changing its proof-selection contract.	Records fixed scope-decision categories separately from GitHub outputs, including invalid-scope handling. Classifies every attempt into bounded event, decision, outcome, and cache labels. Adds an unconditional finalizer that emits atomic JSONL run/cache counters and optional proof duration. Uploads one attempt-specific artifact with 30-day retention and visible failure on missing output.	.github/workflows/ci.yml scripts/kani_proof_scope.py scripts/kani_smoke_metrics.py
Defines and documents the metrics schema, operational semantics, and delivery limitations.	Documents exact metric names, label keys, vocabularies, outcome precedence, cache semantics, timing fallback behavior, and aggregation examples. Records cancellation, timeout, runner-loss, checkout, and setup limitations for finalization and upload.	docs/adr-039-change-scoped-kani-gate.md docs/developers-guide.md
Expands workflow contracts and tests to enforce metrics behavior and wiring.	Covers decision categories, malformed inputs, bounded labels, outcome precedence, cache classification, duration handling, atomic output, and failure warnings. Verifies step ordering, guards, environment paths, finalizer inputs, pinned dependencies, artifact naming, retention, and upload behavior.	tests/workflow_ci.rs tests/workflow_contracts/kani_proof_scope_decision_test.py tests/workflow_contracts/kani_proof_scope_property_test.py tests/workflow_contracts/kani_proof_scope_test.py tests/workflow_contracts/kani_smoke_metrics_contract.py tests/workflow_contracts/kani_smoke_metrics_duration_test.py tests/workflow_contracts/kani_smoke_metrics_property_test.py tests/workflow_contracts/kani_smoke_metrics_test.py tests/workflow_contracts/kani_smoke_metrics_wiring_test.py tests/workflow_contracts/kani_smoke_scope_wiring_test.py
Treats the metrics helper as a proof-scope input for change detection.	Adds the metrics script to the tracked proof-scope paths so changes to observability code trigger Kani proofs.	tools/kani/proof-scope.toml

Assessment against linked issues

Issue	Objective	Addressed	Explanation
#847	Record retained metrics for each Kani workflow attempt, including bounded proof decision and outcome categories, proof duration, and reliable cache hit/miss information across successful, skipped, unreadable-diff, invalid-scope, and proof-failure paths.	✅
#847	Ensure metrics are finalized and retained after normal workflow failures, with an unconditional finalizer, attempt-specific artifact upload, explicit 30-day retention, and documented cancellation, timeout, and delivery limitations.	✅
#847	Keep metric labels bounded and free of paths, IDs, URLs, and free-form errors, while adding behavioral tests and workflow-contract coverage for emission, failure handling, duration, delivery, retention, wiring, and label cardinality.	✅

Possibly linked issues

• Add retained, bounded metrics for kani-smoke #847: PR directly implements issue Add retained, bounded metrics for kani-smoke #847, including bounded metrics, unconditional finalization, 30-day artifacts, documentation, and tests.

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.