[Preset]: Add Parallel Autonomous Run Governance

[hindermath]

Preset ID

parallel-autonomous-run-governance

Preset Name

Parallel Autonomous Run Governance

Version

0.2.0

Description

Coordinates isolated autonomous Spec Kit campaigns with bounded concurrency, mixed agents, resumable consolidation, and governed post-merge closeout.

Author

Thorsten Hindermann

Repository URL

https://github.com/hindermath/spec-kit-preset-parallel-autonomous-run-governance

Download URL

https://github.com/hindermath/spec-kit-preset-parallel-autonomous-run-governance/archive/refs/tags/v0.2.0.zip

Documentation URL

https://github.com/hindermath/spec-kit-preset-parallel-autonomous-run-governance/blob/v0.2.0/README.md

License

MIT

Required Spec Kit Version

=0.8.3

Required Extensions (optional)

None

Templates Provided

• parallel-campaign-template.json - portable schema 1.1 campaign manifest with schema 1.0 compatibility
• parallel-runner-profiles-template.json - local executable and argument-array runner bindings
• parallel-runner-profiles-examples.json - agent-neutral profile examples
• parallel-campaign-state-template.json - aggregate resumable campaign state
• parallel-worker-result-template.json - machine-readable worker result contract
• parallel-autonomous-runbook-template.md - campaign runbook and permission boundaries
• parallel-autonomous-evidence-template.md - campaign evidence and consolidation ledger
• parallel-autonomous-retrospective-template.md - portable campaign retrospective
• agent-file-addendum.md - agent-neutral parallel campaign guidance

Commands Provided

• speckit.parallel-autonomous.md - validate and execute a bounded campaign
• speckit.parallel-autonomous-status.md - inspect campaign and worker state in accessible text or JSON
• speckit.parallel-autonomous-stop.md - request a cooperative stop
• speckit.parallel-autonomous-resume.md - resume after state and worker revalidation
• speckit.parallel-autonomous-consolidate.md - consolidate exact reviewed results under explicit authority

Number of Scripts (optional)

2

Tags

parallel, autonomous, governance, orchestration, resume

Key Features

• Bounded concurrency with a field-validated maximum of three active workers
• Per-worker runner profiles with campaign fallback and agent-neutral model selection
• Cooperative stop, status, resume, and repeated-resume handling
• Provider preflight for exact PR head, draft state, mergeability, reviews, threads, and checks
• Resumable partial consolidation for direct and stacked pull requests
• Manifest-bound post-merge actions and final validation before Completed
• Durable state events, attempt counters, accessible text status, and secret-free JSON status
• Field evidence from 24 workers across six memory-safe language repositories

Testing Checklist

• Preset installs successfully via specify preset add
• Template resolution works correctly after installation
• Documentation is complete and accurate
• Tested on at least one real project

Submission Requirements

• Valid preset.yml manifest included
• Linked README explains how to use this preset and includes the exact tagged specify preset add --from ... command
• LICENSE file included
• GitHub release created with version tag
• Preset ID follows naming conventions

[hindermath]

Hi @mnriem,

I have now published and submitted the new Parallel Autonomous Run Governance preset as v0.2.0. Thank you again for your previous help with the preset catalog.

This release is based on a completed native development field test with 24 workers across six memory-safe language repositories: C#, Go, Java, Python, Rust, and Swift. The campaign used a measured maximum concurrency of three and explicitly exercised stop, status, resume, and a second resume attempt.

The main portable findings are included in the release:

• exact pull-request head and current-review/thread gates before consolidation;
• resumable consolidation after partial merges or a coordinator interruption;
• revalidation of direct and stacked pull-request bases after each merge;
• manifest-bound post-merge closeout before a campaign can become Completed;
• per-worker agent profiles with a campaign-level fallback;
• agent-neutral model selection without prescribed vendor or model names;
• durable attempt and state-event history plus accessible text and secret-free JSON status output.

All 24 worker pull requests and the six repository closeouts were completed. The exact tagged ZIP was then validated in an isolated repository together with the full eight-preset governance matrix, including list, info, resolve, remove/reinstall, and idempotency checks.

I cannot apply the preset-submission label with my current repository permissions. Could you please add that label so the catalog workflow can process the submission?

Many thanks,
Thorsten

[hindermath]

Hi @mnriem,

one final friendly update to this submission: the fleet review and rollout produced a small follow-up release, parallel-autonomous-run-governance v0.2.1.

The tagged release keeps the v0.2.0 contracts and adds the review-derived documentation fixes, including a repository-root-correct help path and bilingual, German-first/English-second native macOS smoke evidence. The published archive was revalidated with SHA-256 1322ef523f2288076237d477eb45656cf330ff3573bf5fcfe3c94e9c204f49d2.

The rollout is now complete:

• the standard eight-preset matrix is merged into Level 0, all seven registered Level-1 repositories, and all 25 registered Level-2 repositories;
• all 33 exact PR heads were verified before merge, and all Copilot review threads were answered and resolved;
• every merged main/default branch was pulled locally and passed the eight-preset matrix check;
• current documentation that previously described six or seven presets was migrated to eight, while historical release and statistics statements were intentionally preserved;
• portable field and fleet findings were also incorporated into patch releases of the other seven governance presets, including resumable consolidation, exact-head and current-review gates, accessible status output, generated-versus-unrelated change handling, and provider-failure classification;
• local parser, matrix, drift, idempotency, and secret checks passed across the fleet. GitHub jobs that still received the known zero-step billing/provider rejection were kept explicitly as N/A, never reported as technical passes.

Thank you for taking a look. The issue still has no preset-submission label on my side; could you please add it when appropriate so the catalog workflow can pick up the current v0.2.1 release?

Many thanks,
Thorsten