[chaos-test] [Chaos R86] null-navigator: staged-subset+two-commits

[github-actions]

Chaos Test PR — Run 86

Persona: null-navigator
Strategy: staged-subset+two-commits
Run: 86

This pull request is an automated chaos test for safe-output create-pull-request bundling.

Scenario Details

• Commit 1: Staged subset — only null-navigator-r86.md added
• Commit 2: Added null-navigator-r86-extra.md + updated first file
• Files: scratchpad/chaos/ scope

Under-tested strategy: staged-subset+two-commits (count: 6 → 7)

Generated by Chaos PR Bundle Fuzzer · sonnet46 1.5M · ◷

• expires on May 30, 2026, 1:22 PM UTC

[github-actions]

Hey there 👋 — thanks for this automated chaos test run (Run 86, null-navigator persona, staged-subset+two-commits strategy)! The PR is well-scoped to the scratchpad/chaos/ area and the description clearly documents the scenario.

One note for the fuzzer workflow itself:

• No test file changes detected — the diff only adds markdown fixture files under scratchpad/chaos/. If the chaos fuzzer is intended to validate PR bundling behaviour, consider whether the workflow should also assert/verify expected outcomes (e.g. a test or assertion file alongside the fixtures) so there's a machine-readable signal of pass/fail baked into the PR diff itself.

If you'd like a hand automating that, here's a ready-to-use prompt for your coding agent:

In the Chaos PR Bundle Fuzzer workflow, after creating the chaos PR fixture files under scratchpad/chaos/, add a small assertion or results file (e.g. scratchpad/chaos/null-navigator-r86-result.json) that records the expected commit count, staged-file list, and strategy name. This gives reviewers and automated checks a machine-readable artefact to validate against, rather than relying solely on the markdown description.

Generated by ✅ Contribution Check · sonnet46 2.6M · ◷