tsl-devkit/docs/superpowers/plans/2026-05-24-playbook-strong-alignment.md

Playbook Strong Alignment Implementation Plan

For agentic workers: REQUIRED SUB-SKILL: Use superpowers:executing-plans or equivalent plan executor. Keep verification local and update memory-bank/progress.md through playbook main-loop commands.

Goal: Update the vendored playbook snapshot, align playbook.toml, and migrate repo-owned agent docs to the latest playbook architecture without touching product README files.

Architecture: Treat docs/standards/playbook/ as a pure vendored subtree, regenerate playbook-managed files from the synced snapshot, then rewrite only the project-owned context files so they match the new memory-bank/ and docs/superpowers/ structure.

Tech Stack: Git subtree, Python playbook scripts, Markdown, repo-local .agents/ rules

---

Plan Meta

• Plan Group: playbook-alignment
• Parent Plan: none
• Verification Scope: local
• Verification Gate: must-pass

Task 1: Update vendored playbook and resync managed files

Files:

• Modify: docs/standards/playbook/**

• Modify: playbook.toml

• Modify: AGENTS.md

• Modify: AGENT_RULES.md

• Modify: docs/prompts/**

• Modify: .agents/**

• Pull the latest upstream playbook into docs/standards/playbook/ with git subtree --squash

• Align playbook.toml to the new config shape and add claude_md = "CLAUDE.md"

• Run python docs/standards/playbook/scripts/playbook.py -config playbook.toml to resync managed files

Task 2: Migrate project-owned context to the new architecture

Files:

• Modify: AGENT_RULES.local.md

• Modify: memory-bank/project-brief.md

• Modify: memory-bank/tech-context.md

• Modify: memory-bank/system-patterns.md

• Modify: memory-bank/active-context.md

• Modify: memory-bank/progress.md

• Modify: memory-bank/decisions.md

• Rewrite AGENT_RULES.local.md so it keeps only tsl-devkit-specific rules

• Migrate memory-bank to project-brief / tech-context / system-patterns / active-context / progress / decisions

• Preserve repo-specific facts such as layer boundaries, Tree-sitter synchronization and CIFS/git caveats

Task 3: Remove old architecture remnants and move the plan entry point

Files:

• Delete: memory-bank/architecture.md

• Delete: memory-bank/tech-stack.md

• Delete: docs/prompts/coding/review.md

• Delete: docs/prompts/meta/prompt-generator.md

• Deprecate: docs/plans/2026-05-24-playbook-strong-alignment.md

• Create: docs/superpowers/plans/2026-05-24-playbook-strong-alignment.md

• Create the new plan entry point under docs/superpowers/plans/

• Delete prompt and memory-bank files that are no longer part of the latest playbook layout

• Keep product README files out of scope for this migration

• Replace the old docs/plans/ file with a deprecation stub because the current CIFS mount rejected path migration

Task 4: Re-record workflow state, verify, and create the final commit

Files:

• Modify: memory-bank/progress.md

• Commit: aligned playbook snapshot, rules, prompts, memory-bank and plan files

• Re-record the plan with python docs/standards/playbook/scripts/playbook.py -record-plan ...

• Use main_loop.py claim / finish to update authoritative workflow blocks

• Re-run sync and diff checks to verify idempotence and whitespace cleanliness

• Commit with repository commit-message convention