bmad-plus 0.17.0 → 0.18.0

package/tools/build/generated-adapters/CLAUDE.md

@@ -1,135 +0,0 @@
-<!-- AUTO-GENERATED by tools/build/generate-adapters.js — DO NOT EDIT. -->
-<!-- Source of truth: registry.yaml → spine AGENTS.md. Regenerate: node tools/build/generate-adapters.js -->
-
-# BMAD+ — Adapter for claude-code
-
-This file is a THIN adapter. The single source of truth for agents, packs,
-skills, and workflows is the spine file: **AGENTS.md** (project root).
-Read AGENTS.md first; this adapter only adds tool-specific invocation notes.
-
-## Tool notes — claude-code
-
-- Claude Code loads `CLAUDE.md` from the project root automatically.
-- Activate an agent by saying its name (see the spine for the full roster).
-- Skills are loaded from `.agents/skills/`; auto-activation triggers live in `.agents/data/role-triggers.yaml`.
-
-## Project Context
-
-This project uses BMAD+, an augmented AI-driven development framework.
-Based on BMAD-METHOD v6.6.0 with multi-role agents, autopilot mode, and parallel execution.
-
-<!-- Hand-authored project instructions — single source: tools/build/adapters.config.js. -->
-<!-- Edit the config, then regenerate; never edit this file directly. -->
-
-## Agents
-
-To activate an agent, say its name or persona:
-
-- **Atlas** (Strategist) — Business analysis + Product management
-- **Forge** (Architect-Dev) — Architecture + Development + Documentation
-- **Sentinel** (Quality) — QA + UX review
-- **Nexus** (Orchestrator) — Sprint management + Autopilot + Parallel execution
-- **Shadow** (OSINT) — Investigation + Scraping + Psychoprofiling (OSINT)
-- **Maker** (Agent Creator) — Design, build, validate, and package new BMAD+ agents (Maker)
-- **Shield** (GRC) — Compliance agents for GDPR, ISO 27001, SOC 2, HIPAA, EU AI Act, DORA, NIS2 and more (Shield)
-- **SEO Scout** (Technical Scanner) — Crawling + technical inspection + performance (SEO)
-- **SEO Chief** (Strategist & Reporter) — Scoring + strategy + reporting (SEO)
-- **SEO Judge** (Content & AI Analyst) — Content quality + structured data + GEO analysis (SEO)
-- **Zecher** (זכר, Memory Guardian) — Memory Archivist — Persistent cross-session memory, consolidation, project scanning, context recall, session handoffs (Memory)
-- **Miriam** (מרים) — Business Analyst — Strategic analysis, research, product briefs (Dev Studio)
-- **Huldah** (חולדה) — Technical Writer — Documentation, diagrams, editorial review (Dev Studio)
-- **Yosef** (יוסף) — Product Manager — PRD, requirements, feature prioritization (Dev Studio)
-- **Rachel** (רחל) — UX Designer — User experience, wireframes, empathy mapping (Dev Studio)
-- **Bezalel** (בצלאל) — System Architect — Architecture, ADRs, epics & stories (Dev Studio)
-- **Oholiab** (אהליאב) — Senior Engineer — TDD, sprint, code review, implementation (Dev Studio)
-- **Backup Manager** (Backup & Restore) — Timestamped backups, restoration, and rotation (Backup)
-- **Animated Website Creator** (Website Builder) — Video-driven scrolling websites (Animated)
-
-## Skills
-
-- Installed projects load skills from `.agents/skills/`; in this repository the sources live in `src/bmad-plus/skills/` and `src/bmad-plus/agents/`.
-- Each agent has a SKILL.md with capabilities, activation protocol, and role-switching rules.
-- Auto-activation triggers: `.agents/data/role-triggers.yaml` (source: `src/bmad-plus/data/role-triggers.yaml`).
-
-## Project Structure
-
-- `src/bmad-plus/` — Custom module (agents, skills, data)
-- `monitor/` — Upstream monitoring system (VPS)
-- `mcp-server/` — Audit 360° MCP Server
-- `osint-agent-package/` — OSINT package
-- `upstream/` — BMAD-METHOD reference clone
-
-## Communication
-
-- User name: laurent
-- Default language: Français for user-facing content, English for code and technical docs.
-
-## Commit Rules
-
-- NEVER add "Co-Authored-By: Claude" or any AI co-author attribution.
-- The sole author is Laurent Rochetta.
-
-## Repository Maintenance Rule
-
-When updating the main README.md (English), you MUST synchronously update all translations in the readme-international/ directory (fr, es, de).
-
-## Memory Protocol (Karpathy Guardrails)
-
-Agents MUST follow these behavioral principles:
-
-### G1 — Think Before Coding
-
-- State assumptions explicitly. If uncertain, ask.
-- Check `.agents/memory/decisions.md` for prior decisions before re-deciding.
-
-### G2 — Simplicity First
-
-- Minimum code that solves the problem. Nothing speculative.
-- Check `.agents/memory/patterns.md` for existing solutions.
-
-### G3 — Surgical Changes
-
-- Touch only what you must. Match existing style.
-- Log surprises in `.agents/memory/lessons.md`.
-
-### G4 — Goal-Driven Execution
-
-- Define success criteria before implementing.
-- Log non-obvious decisions in `.agents/memory/decisions.md`.
-
-### Memory Files
-
-- `.agents/memory/decisions.md` — Read at session start, write when making decisions
-- `.agents/memory/lessons.md` — Write when something unexpected happens
-- `.agents/memory/patterns.md` — Write when a reusable pattern is validated
-- `.agents/memory/context.md` — Update at session end with project state
-
-## Framework updates
-
-Once per session, check the installed framework before starting project work. If this is the BMAD+ source checkout, use `node tools/cli/bmad-plus-cli.js update-check --json`; otherwise use `npx --no-install bmad-plus update-check --json` from the project root.
-If the command is unavailable or the result is off, unknown, offline or stale, continue the user task. Do not install a CLI just to perform this check or claim that an offline version is current.
-When `updateAvailable` is true and `canAutoApply` is true, run the same CLI with `update --latest --auto`. This follows the project update policy; it does not grant permission to expand that policy.
-When an update is available but automatic application is not eligible, explain the version and reason. An explicit request to update can be fulfilled with `update --latest --yes`; do not change the policy, overwrite customized files, or resolve conflicts without the user request covering that action.
-After a successful update, reread the project spine and relevant installed skills, or start a fresh session if the client cannot reload them. Respect preserved local instructions. Check only once even when several adapters include this section.
-
-## Registry facts (computed from registry.yaml — never hand-typed)
-
-- Product: BMAD+ v0.17.0 (derived from BMAD-METHOD v6.6.0)
-- Models supported: claude, gpt, gemini, local (model-agnostic by contract)
-- Packs (9): Core, OSINT, Maker, Shield, SEO, Memory, Dev Studio, Backup, Animated
-- Installer agents (14 across all packs):
-  - Core (required): 4 installer agents — Core agents & skills
-  - OSINT: 1 installer agent — OSINT & investigation — 2 compliance frameworks
-  - Maker: 1 installer agent — Agent creation toolkit
-  - Shield: 1 installer agent — GRC compliance (26 frameworks) — 27 specialized agents, 11 workflows, 26 compliance frameworks
-  - SEO: 3 installer agents — SEO audit & optimization — 1 compliance frameworks
-  - Memory: 1 installer agent — Persistent cross-session memory
-  - Dev Studio: 1 installer agent — SDLC automation (6 agents, specialized workflows) — 6 sub-agents, 38 workflows
-  - Backup: 1 installer agent — Backup & restore
-  - Animated: 1 installer agent — Animated website agents
-
-## Key commands
-
-- `bmad-help` — Show all available agents and skills
-- `autopilot` — Launch the orchestrator (Nexus) in full pipeline mode
-- `parallel` — Enable parallel multi-agent execution

package/tools/build/generated-adapters/CONVENTIONS.md

@@ -1,134 +0,0 @@
-<!-- AUTO-GENERATED by tools/build/generate-adapters.js — DO NOT EDIT. -->
-<!-- Source of truth: registry.yaml → spine AGENTS.md. Regenerate: node tools/build/generate-adapters.js -->
-
-# BMAD+ — Adapter for aider
-
-This file is a THIN adapter. The single source of truth for agents, packs,
-skills, and workflows is the spine file: **AGENTS.md** (project root).
-Read AGENTS.md first; this adapter only adds tool-specific invocation notes.
-
-## Tool notes — aider
-
-- Aider does not auto-load convention files: add `read: CONVENTIONS.md` to `.aider.conf.yml` or run `/read CONVENTIONS.md`.
-- Treat the spine as the single source of truth; this file only adds Aider-specific notes.
-
-## Project Context
-
-This project uses BMAD+, an augmented AI-driven development framework.
-Based on BMAD-METHOD v6.6.0 with multi-role agents, autopilot mode, and parallel execution.
-
-<!-- Hand-authored project instructions — single source: tools/build/adapters.config.js. -->
-<!-- Edit the config, then regenerate; never edit this file directly. -->
-
-## Agents
-
-To activate an agent, say its name or persona:
-
-- **Atlas** (Strategist) — Business analysis + Product management
-- **Forge** (Architect-Dev) — Architecture + Development + Documentation
-- **Sentinel** (Quality) — QA + UX review
-- **Nexus** (Orchestrator) — Sprint management + Autopilot + Parallel execution
-- **Shadow** (OSINT) — Investigation + Scraping + Psychoprofiling (OSINT)
-- **Maker** (Agent Creator) — Design, build, validate, and package new BMAD+ agents (Maker)
-- **Shield** (GRC) — Compliance agents for GDPR, ISO 27001, SOC 2, HIPAA, EU AI Act, DORA, NIS2 and more (Shield)
-- **SEO Scout** (Technical Scanner) — Crawling + technical inspection + performance (SEO)
-- **SEO Chief** (Strategist & Reporter) — Scoring + strategy + reporting (SEO)
-- **SEO Judge** (Content & AI Analyst) — Content quality + structured data + GEO analysis (SEO)
-- **Zecher** (זכר, Memory Guardian) — Memory Archivist — Persistent cross-session memory, consolidation, project scanning, context recall, session handoffs (Memory)
-- **Miriam** (מרים) — Business Analyst — Strategic analysis, research, product briefs (Dev Studio)
-- **Huldah** (חולדה) — Technical Writer — Documentation, diagrams, editorial review (Dev Studio)
-- **Yosef** (יוסף) — Product Manager — PRD, requirements, feature prioritization (Dev Studio)
-- **Rachel** (רחל) — UX Designer — User experience, wireframes, empathy mapping (Dev Studio)
-- **Bezalel** (בצלאל) — System Architect — Architecture, ADRs, epics & stories (Dev Studio)
-- **Oholiab** (אהליאב) — Senior Engineer — TDD, sprint, code review, implementation (Dev Studio)
-- **Backup Manager** (Backup & Restore) — Timestamped backups, restoration, and rotation (Backup)
-- **Animated Website Creator** (Website Builder) — Video-driven scrolling websites (Animated)
-
-## Skills
-
-- Installed projects load skills from `.agents/skills/`; in this repository the sources live in `src/bmad-plus/skills/` and `src/bmad-plus/agents/`.
-- Each agent has a SKILL.md with capabilities, activation protocol, and role-switching rules.
-- Auto-activation triggers: `.agents/data/role-triggers.yaml` (source: `src/bmad-plus/data/role-triggers.yaml`).
-
-## Project Structure
-
-- `src/bmad-plus/` — Custom module (agents, skills, data)
-- `monitor/` — Upstream monitoring system (VPS)
-- `mcp-server/` — Audit 360° MCP Server
-- `osint-agent-package/` — OSINT package
-- `upstream/` — BMAD-METHOD reference clone
-
-## Communication
-
-- User name: laurent
-- Default language: Français for user-facing content, English for code and technical docs.
-
-## Commit Rules
-
-- NEVER add "Co-Authored-By: Claude" or any AI co-author attribution.
-- The sole author is Laurent Rochetta.
-
-## Repository Maintenance Rule
-
-When updating the main README.md (English), you MUST synchronously update all translations in the readme-international/ directory (fr, es, de).
-
-## Memory Protocol (Karpathy Guardrails)
-
-Agents MUST follow these behavioral principles:
-
-### G1 — Think Before Coding
-
-- State assumptions explicitly. If uncertain, ask.
-- Check `.agents/memory/decisions.md` for prior decisions before re-deciding.
-
-### G2 — Simplicity First
-
-- Minimum code that solves the problem. Nothing speculative.
-- Check `.agents/memory/patterns.md` for existing solutions.
-
-### G3 — Surgical Changes
-
-- Touch only what you must. Match existing style.
-- Log surprises in `.agents/memory/lessons.md`.
-
-### G4 — Goal-Driven Execution
-
-- Define success criteria before implementing.
-- Log non-obvious decisions in `.agents/memory/decisions.md`.
-
-### Memory Files
-
-- `.agents/memory/decisions.md` — Read at session start, write when making decisions
-- `.agents/memory/lessons.md` — Write when something unexpected happens
-- `.agents/memory/patterns.md` — Write when a reusable pattern is validated
-- `.agents/memory/context.md` — Update at session end with project state
-
-## Framework updates
-
-Once per session, check the installed framework before starting project work. If this is the BMAD+ source checkout, use `node tools/cli/bmad-plus-cli.js update-check --json`; otherwise use `npx --no-install bmad-plus update-check --json` from the project root.
-If the command is unavailable or the result is off, unknown, offline or stale, continue the user task. Do not install a CLI just to perform this check or claim that an offline version is current.
-When `updateAvailable` is true and `canAutoApply` is true, run the same CLI with `update --latest --auto`. This follows the project update policy; it does not grant permission to expand that policy.
-When an update is available but automatic application is not eligible, explain the version and reason. An explicit request to update can be fulfilled with `update --latest --yes`; do not change the policy, overwrite customized files, or resolve conflicts without the user request covering that action.
-After a successful update, reread the project spine and relevant installed skills, or start a fresh session if the client cannot reload them. Respect preserved local instructions. Check only once even when several adapters include this section.
-
-## Registry facts (computed from registry.yaml — never hand-typed)
-
-- Product: BMAD+ v0.17.0 (derived from BMAD-METHOD v6.6.0)
-- Models supported: claude, gpt, gemini, local (model-agnostic by contract)
-- Packs (9): Core, OSINT, Maker, Shield, SEO, Memory, Dev Studio, Backup, Animated
-- Installer agents (14 across all packs):
-  - Core (required): 4 installer agents — Core agents & skills
-  - OSINT: 1 installer agent — OSINT & investigation — 2 compliance frameworks
-  - Maker: 1 installer agent — Agent creation toolkit
-  - Shield: 1 installer agent — GRC compliance (26 frameworks) — 27 specialized agents, 11 workflows, 26 compliance frameworks
-  - SEO: 3 installer agents — SEO audit & optimization — 1 compliance frameworks
-  - Memory: 1 installer agent — Persistent cross-session memory
-  - Dev Studio: 1 installer agent — SDLC automation (6 agents, specialized workflows) — 6 sub-agents, 38 workflows
-  - Backup: 1 installer agent — Backup & restore
-  - Animated: 1 installer agent — Animated website agents
-
-## Key commands
-
-- `bmad-help` — Show all available agents and skills
-- `autopilot` — Launch the orchestrator (Nexus) in full pipeline mode
-- `parallel` — Enable parallel multi-agent execution

package/tools/build/generated-adapters/GEMINI.md

@@ -1,139 +0,0 @@
-<!-- AUTO-GENERATED by tools/build/generate-adapters.js — DO NOT EDIT. -->
-<!-- Source of truth: registry.yaml → spine AGENTS.md. Regenerate: node tools/build/generate-adapters.js -->
-
-# BMAD+ — Adapter for gemini-cli + antigravity
-
-This file is a THIN adapter. The single source of truth for agents, packs,
-skills, and workflows is the spine file: **AGENTS.md** (project root).
-Read AGENTS.md first; this adapter only adds tool-specific invocation notes.
-
-## Tool notes — gemini-cli
-
-- Gemini CLI loads `GEMINI.md` from the project root automatically.
-- Activate an agent by saying its name (see the spine for the full roster).
-
-## Tool notes — antigravity
-
-- Antigravity reads the same `GEMINI.md` file (detected via `.gemini/antigravity`).
-- Concurrency rule: coordinate with other agentic CLIs before writing — never edit the same files simultaneously.
-
-## Project Context
-
-This project uses BMAD+, an augmented AI-driven development framework.
-Based on BMAD-METHOD v6.6.0 with multi-role agents, autopilot mode, and parallel execution.
-
-<!-- Hand-authored project instructions — single source: tools/build/adapters.config.js. -->
-<!-- Edit the config, then regenerate; never edit this file directly. -->
-
-## Agents
-
-To activate an agent, say its name or persona:
-
-- **Atlas** (Strategist) — Business analysis + Product management
-- **Forge** (Architect-Dev) — Architecture + Development + Documentation
-- **Sentinel** (Quality) — QA + UX review
-- **Nexus** (Orchestrator) — Sprint management + Autopilot + Parallel execution
-- **Shadow** (OSINT) — Investigation + Scraping + Psychoprofiling (OSINT)
-- **Maker** (Agent Creator) — Design, build, validate, and package new BMAD+ agents (Maker)
-- **Shield** (GRC) — Compliance agents for GDPR, ISO 27001, SOC 2, HIPAA, EU AI Act, DORA, NIS2 and more (Shield)
-- **SEO Scout** (Technical Scanner) — Crawling + technical inspection + performance (SEO)
-- **SEO Chief** (Strategist & Reporter) — Scoring + strategy + reporting (SEO)
-- **SEO Judge** (Content & AI Analyst) — Content quality + structured data + GEO analysis (SEO)
-- **Zecher** (זכר, Memory Guardian) — Memory Archivist — Persistent cross-session memory, consolidation, project scanning, context recall, session handoffs (Memory)
-- **Miriam** (מרים) — Business Analyst — Strategic analysis, research, product briefs (Dev Studio)
-- **Huldah** (חולדה) — Technical Writer — Documentation, diagrams, editorial review (Dev Studio)
-- **Yosef** (יוסף) — Product Manager — PRD, requirements, feature prioritization (Dev Studio)
-- **Rachel** (רחל) — UX Designer — User experience, wireframes, empathy mapping (Dev Studio)
-- **Bezalel** (בצלאל) — System Architect — Architecture, ADRs, epics & stories (Dev Studio)
-- **Oholiab** (אהליאב) — Senior Engineer — TDD, sprint, code review, implementation (Dev Studio)
-- **Backup Manager** (Backup & Restore) — Timestamped backups, restoration, and rotation (Backup)
-- **Animated Website Creator** (Website Builder) — Video-driven scrolling websites (Animated)
-
-## Skills
-
-- Installed projects load skills from `.agents/skills/`; in this repository the sources live in `src/bmad-plus/skills/` and `src/bmad-plus/agents/`.
-- Each agent has a SKILL.md with capabilities, activation protocol, and role-switching rules.
-- Auto-activation triggers: `.agents/data/role-triggers.yaml` (source: `src/bmad-plus/data/role-triggers.yaml`).
-
-## Project Structure
-
-- `src/bmad-plus/` — Custom module (agents, skills, data)
-- `monitor/` — Upstream monitoring system (VPS)
-- `mcp-server/` — Audit 360° MCP Server
-- `osint-agent-package/` — OSINT package
-- `upstream/` — BMAD-METHOD reference clone
-
-## Communication
-
-- User name: laurent
-- Default language: Français for user-facing content, English for code and technical docs.
-
-## Commit Rules
-
-- NEVER add "Co-Authored-By: Claude" or any AI co-author attribution.
-- The sole author is Laurent Rochetta.
-
-## Repository Maintenance Rule
-
-When updating the main README.md (English), you MUST synchronously update all translations in the readme-international/ directory (fr, es, de).
-
-## Memory Protocol (Karpathy Guardrails)
-
-Agents MUST follow these behavioral principles:
-
-### G1 — Think Before Coding
-
-- State assumptions explicitly. If uncertain, ask.
-- Check `.agents/memory/decisions.md` for prior decisions before re-deciding.
-
-### G2 — Simplicity First
-
-- Minimum code that solves the problem. Nothing speculative.
-- Check `.agents/memory/patterns.md` for existing solutions.
-
-### G3 — Surgical Changes
-
-- Touch only what you must. Match existing style.
-- Log surprises in `.agents/memory/lessons.md`.
-
-### G4 — Goal-Driven Execution
-
-- Define success criteria before implementing.
-- Log non-obvious decisions in `.agents/memory/decisions.md`.
-
-### Memory Files
-
-- `.agents/memory/decisions.md` — Read at session start, write when making decisions
-- `.agents/memory/lessons.md` — Write when something unexpected happens
-- `.agents/memory/patterns.md` — Write when a reusable pattern is validated
-- `.agents/memory/context.md` — Update at session end with project state
-
-## Framework updates
-
-Once per session, check the installed framework before starting project work. If this is the BMAD+ source checkout, use `node tools/cli/bmad-plus-cli.js update-check --json`; otherwise use `npx --no-install bmad-plus update-check --json` from the project root.
-If the command is unavailable or the result is off, unknown, offline or stale, continue the user task. Do not install a CLI just to perform this check or claim that an offline version is current.
-When `updateAvailable` is true and `canAutoApply` is true, run the same CLI with `update --latest --auto`. This follows the project update policy; it does not grant permission to expand that policy.
-When an update is available but automatic application is not eligible, explain the version and reason. An explicit request to update can be fulfilled with `update --latest --yes`; do not change the policy, overwrite customized files, or resolve conflicts without the user request covering that action.
-After a successful update, reread the project spine and relevant installed skills, or start a fresh session if the client cannot reload them. Respect preserved local instructions. Check only once even when several adapters include this section.
-
-## Registry facts (computed from registry.yaml — never hand-typed)
-
-- Product: BMAD+ v0.17.0 (derived from BMAD-METHOD v6.6.0)
-- Models supported: claude, gpt, gemini, local (model-agnostic by contract)
-- Packs (9): Core, OSINT, Maker, Shield, SEO, Memory, Dev Studio, Backup, Animated
-- Installer agents (14 across all packs):
-  - Core (required): 4 installer agents — Core agents & skills
-  - OSINT: 1 installer agent — OSINT & investigation — 2 compliance frameworks
-  - Maker: 1 installer agent — Agent creation toolkit
-  - Shield: 1 installer agent — GRC compliance (26 frameworks) — 27 specialized agents, 11 workflows, 26 compliance frameworks
-  - SEO: 3 installer agents — SEO audit & optimization — 1 compliance frameworks
-  - Memory: 1 installer agent — Persistent cross-session memory
-  - Dev Studio: 1 installer agent — SDLC automation (6 agents, specialized workflows) — 6 sub-agents, 38 workflows
-  - Backup: 1 installer agent — Backup & restore
-  - Animated: 1 installer agent — Animated website agents
-
-## Key commands
-
-- `bmad-help` — Show all available agents and skills
-- `autopilot` — Launch the orchestrator (Nexus) in full pipeline mode
-- `parallel` — Enable parallel multi-agent execution

package/tools/build/generated-adapters/README.md

@@ -1,79 +0,0 @@
-# Generated Multi-CLI Adapters (Pillar 5 — portability lock)
-
-Everything in this directory **except this README** is generated by:
-
-```
-node tools/build/generate-adapters.js
-```
-
-from the root `registry.yaml` (`targets.adapters[]` + pack declarations) plus
-the hand-authored sections in `tools/build/adapters.config.js`.
-
-## ADOPTED — the root files are now generated too
-
-Since v0.11.0 the generated adapters are **load-bearing**: the live repo-root
-`AGENTS.md` (spine), `CLAUDE.md`, `GEMINI.md`, `.cursor/rules/bmad-plus.mdc`,
-`.codex/AGENTS.md`, `.opencode/AGENTS.md` and `CONVENTIONS.md` are written by
-
-```
-node tools/build/generate-adapters.js --adopt
-```
-
-and drift-gated by
-
-```
-node tools/build/generate-adapters.js --check --adopt
-```
-
-(also enforced by `tests/unit/generate-adapters.test.js`). This directory
-remains as a PREVIEW target so you can inspect what a registry/config change
-produces before adopting it (`git diff` on this dir vs the root files).
-
-## Where content comes from
-
-- **Derived facts** (product/version, pack list, every count) — `registry.yaml`.
-  No human ever types a count, so adapters cannot drift or lie.
-- **Hand-authored project instructions** (agent roster, skills paths, project
-  structure, communication, commit rules, README i18n rule, Karpathy memory
-  protocol) — `tools/build/adapters.config.js`. The pre-adoption root
-  CLAUDE.md/GEMINI.md/AGENTS.md content was folded there verbatim (union of
-  all three, nothing dropped).
-- **Tool notes** — `TOOL_NOTES` in `tools/build/generate-adapters.js`.
-
-## How to change the root configs
-
-1. Edit `tools/build/adapters.config.js` (instructions) or `registry.yaml`
-   (facts/targets) — NEVER the generated files themselves.
-2. Regenerate: `node tools/build/generate-adapters.js --adopt` (previews:
-   run without `--adopt`).
-3. `node tools/build/generate-adapters.js --check --adopt` must pass.
-
-A safety guard refuses `--adopt` over a root file that lacks the
-`AUTO-GENERATED` marker (i.e. hand-written content that was never folded in);
-override only with `--force` after folding the content into the config.
-
-## What each file is
-
-One generated file per unique `targets.adapters[].file` entry, plus the spine
-(`targets.spine`). Tools that share a file (e.g. `gemini-cli` + `antigravity`
-→ `GEMINI.md`) are grouped into a single adapter with one "Tool notes" block
-per tool. Every adapter:
-
-- points the tool at the **spine** (`targets.spine`, currently `AGENTS.md`) as
-  the single source of truth,
-- adds tool-specific invocation notes,
-- inlines the full hand-authored project instructions (most tools auto-load
-  only their own file, so a thin pointer would silently lose behavior),
-- lists pack/agent facts **computed** from `registry.yaml`.
-
-## Drift check
-
-```
-node tools/build/generate-adapters.js --check          # this preview dir
-node tools/build/generate-adapters.js --check --adopt  # the live root files
-```
-
-Each exits 1 if any file no longer matches what the registry + config generate
-(same contract as `tools/build/generate.js --check` for `packs.js`). Both are
-suitable for CI. Comparisons are EOL-insensitive so Windows checkouts do not
-false-positive.