Skip to content

policy: one bundled H6 plan choice for new projects - #32

Merged
phr-devil merged 11 commits into
mainfrom
feat/bundled-h6-choice-20261116
Nov 16, 2026
Merged

phr-devil merged 11 commits into
mainfrom
feat/bundled-h6-choice-20261116

Conversation

@phr-devil

Copy link
Copy Markdown
Owner

Why

New-project startup prompts already supply scope/goals (H1/H2). Existing H1-H6 rules can cause repeated questions or agent-authored declarations with no true author choice.

Changes

  • Keep v1.3 H1-H10 identities/dependencies and historic compatibility. AI first drafts H3-H5; a single specific user reply after the current H6 plan card selects/binds H3-H6 together.
  • Do not let generic 'continue', initial broad exploration delegation, silence or model recommendation count as this reply. Pending H6 blocks dependent new-mainline execution but not independent permitted read-only activity.
  • Clarify behavior across WORKFLOW, AGENTS, START, author strategy review, data discovery, next-action policy, core README and JSON machine mirror. Add a policy regression assertion.

Guardrails

No new operation permissions, project ledger, H numbers or retroactive old-project approvals. Existing advisory validators cannot authenticate chat replies; real hard host enforcement remains separate.

Validation

Existing PD-RW tests via CI. This is policy interaction alignment, not a claim of live WorkBuddy compliance.

@phr-devil
phr-devil merged commit e70a9be into main Nov 16, 2026
0 of 3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant