cfsa-antigravity 2.19.2 → 2.19.4

package/README.md

@@ -10,7 +10,7 @@ A pipeline that turns a raw idea into exhaustively specified, test-driven, produ
 npx cfsa-antigravity init
 ```
 
-This installs the `.agent/` folder, `docs/` structure, and agent config files into your project.
+This installs the selected agent runtime (`.agent/` by default, or `.claude/` with `--agent claude`), the `docs/` structure, and agent config files into your project.
 
 ## CLI
 
@@ -22,13 +22,19 @@ This installs the `.agent/` folder, `docs/` structure, and agent config files in
 | `cfsa-antigravity init --dry-run` | Preview what would be installed |
 | `cfsa-antigravity init --path ./dir` | Install into specific directory |
 
-### ⚠️ Important Note on `.gitignore`
-If you are using AI-powered editors like **Cursor** or **Windsurf**, adding the `.agent/` folder to your `.gitignore` may prevent the IDE from indexing the workflows. This results in slash commands (like `/ideate`, `/create-prd`) not appearing in the chat suggestion dropdown.
+### Choosing the runtime directory
+- Antigravity-style installs use `.agent/`
+- Claude Code installs use `.claude/`
+
+If you are using an editor that indexes agent files for slash commands, do not hide the runtime directory you actually installed.
 
 **Recommended Solution:**
-To keep the `.agent/` folder local (not tracked by Git) while maintaining AI functionality:
-1. Ensure `.agent/` is **NOT** in your project's `.gitignore`.
-2. Instead, add it to your local exclude file: `.git/info/exclude`
+Keep the installed runtime directory out of shared `.gitignore` rules when your tool needs to index it. Prefer `.git/info/exclude` for local-only exclusions.
+
+Examples:
+1. Antigravity install → keep `.agent/` out of `.gitignore`
+2. Claude install → keep `.claude/` out of `.gitignore`
+3. Use `.git/info/exclude` for local exclusions instead of changing shared ignore rules
    
 ## Get Started
 
@@ -49,7 +55,7 @@ The kit evolves independently of your project. To pull improvements into an exis
 This performs a **semantic merge** — it applies new workflows, skills, and rules from the upstream kit while preserving your project-specific values (tech stack, validation commands, filled placeholders). It will never overwrite your project decisions.
 
 - First sync does a full comparison; subsequent syncs are incremental (commit-scoped)
-- Tracks sync state in `.agent/kit-sync.md` so it knows what changed since last update
+- Tracks sync state in the installed agent runtime (`.agent/kit-sync.md` or `.claude/kit-sync.md`) so it knows what changed since last update
 - Flags any structural migrations needed (e.g., ideation format changes)
 
 ## Documentation

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "cfsa-antigravity",
-  "version": "2.19.2",
+  "version": "2.19.4",
   "description": "CFSA Pipeline — Constraint-First Specification Architecture for AI agents. Production-grade from line one.",
   "scripts": {
     "changeset": "changeset",

package/template/.agent/kit-sync.md

@@ -1,6 +1,6 @@
 # Kit Sync State
 
 upstream: https://github.com/RepairYourTech/cfsa-antigravity
-last_synced_commit: 15665e628973ef08a1dae1c96a9567c4170f4f66
-last_synced_at: 2026-04-06T05:36:53Z
-kit_version: 2.19.2
+last_synced_commit: 515f87acc1239e276a85bf46b0d33e8925659a57
+last_synced_at: 2026-04-10T05:15:43Z
+kit_version: 2.19.4

package/template/.claude/README.md

@@ -12,7 +12,7 @@ This directory contains the CFSA (Constraint-First Specification Architecture) p
 │   └── utilities/       # Helper skills (resolution, templates, etc.)
 ├── rules/               # Always-active rules that apply to every task
 ├── instructions/        # Core directives (workflow, tech-stack, patterns)
-├── skill-library/       # Shared skill library (symlinked to ../.agent/skill-library/)
+├── skill-library/       # Claude-owned skill library
 └── memory/              # CFSA-specific memory protocols
     ├── decisions.md     # Locked decisions log
     ├── patterns.md      # Reusable patterns with confidence scores
@@ -20,19 +20,15 @@ This directory contains the CFSA (Constraint-First Specification Architecture) p
     └── sessions/        # Session logs for continuity
 ```
 
-## Parallel Structure with .agent/
+## Parallel Structure
 
-This directory parallels the `.agent/` directory structure used by the Antigravity CLI. Both versions share:
-
-- The same skill library (`skill-library/`)
-- The same documentation (`docs/`)
-- The same core philosophy and pipeline stages
+This directory is the standalone Claude Code runtime for the CFSA pipeline. It sits alongside the Antigravity runtime in this repository, but `.claude/` owns its Claude execution assets locally.
 
 ## Key Differences from Antigravity Version
 
 1. **Skills vs Workflows**: Claude Code uses skills instead of passive workflow markdown files
 2. **Task System**: Uses Claude Code's built-in Tasks system for progress tracking
-3. **Memory Location**: Uses `.claude/memory/` instead of `.agent/progress/`
+3. **Progress + Memory**: Uses `.claude/progress/` for pipeline state and `.claude/memory/` for CFSA memory
 4. **Invocation**: Workflows are invoked as skills rather than slash commands in markdown
 
 ## Installation

package/template/.claude/instructions/structure.md

@@ -20,10 +20,10 @@
 ## Protected Files (Do Not Modify Without Approval)
 - `AGENTS.md` — Project agent config
 - `docs/plans/ENGINEERING-STANDARDS.md` — Quality bar
-- `.agent/instructions/*` — Agent rules
+- `.claude/instructions/*` — Agent rules
 - `package.json` — Dependencies (add carefully)
 - `tsconfig.json` — TypeScript config
 
 ## Notes
-- All agent config lives in `.agent/` (canonical for all agents)
+- Claude Code agent config lives in `.claude/`
 - Frontend and API may be separately deployed depending on architecture

package/template/.claude/instructions/workflow.md

@@ -4,12 +4,12 @@ Before taking any action on a task:
 
 ## 1. Understand Context
 - Read the agent config file at project root (`AGENTS.md` for Antigravity, `GEMINI.md` for Gemini CLI, or equivalent for your agent)
-- Read relevant `.agent/instructions/` files for the task type
+- Read relevant `.claude/instructions/` files for the task type
 - Check [Engineering Standards](../../docs/plans/ENGINEERING-STANDARDS.md) for quality bar — if this file doesn't exist yet, the pipeline hasn't reached /create-prd; run /ideate then /create-prd first
-- **Session Resumption**: If `.agent/progress/index.md` exists, read `.agent/skills/session-continuity/protocols/01-session-resumption.md` and follow the **Session Resumption Protocol** to load cross-session context and identify the resumption point
+- **Session Resumption**: If `.claude/progress/index.md` exists, read `.claude/skills/session-continuity/protocols/01-session-resumption.md` and follow the **Session Resumption Protocol** to load cross-session context and identify the resumption point
 
 ## 2. Check Skills
-- Scan `.agent/skills/` for applicable skills
+- Scan `.claude/skills/` for applicable skills
 - Load relevant skill SKILL.md before proceeding
 - Key skills: `brainstorming` (before creative work), `tdd-workflow` (before features), `clean-code` (during review)
 
@@ -28,7 +28,7 @@ Before taking any action on a task:
 ## 4. Validate (MANDATORY)
 After **every** code change, run:
 ```bash
-See `.agent/instructions/commands.md` for the validation command.
+See `.claude/instructions/commands.md` for the validation command.
 ```
 
 Do NOT mark a task complete until all validations pass.
@@ -37,8 +37,8 @@ Do NOT mark a task complete until all validations pass.
 
 After completing a workflow or substantial task:
 
-- **Pattern Extraction**: Read `.agent/skills/session-continuity/protocols/04-pattern-extraction.md` and follow the **Pattern Extraction Protocol**. Reflect on what worked, what didn't, and log reusable patterns to `memory/patterns.md`. Skip only if the task was trivial (routine, nothing new learned).
-- **Session Close**: Read `.agent/skills/session-continuity/protocols/05-session-close.md` and follow the **Session Close Protocol**. Write a session log to `.agent/progress/sessions/` so the next session can resume cleanly.
+- **Pattern Extraction**: Read `.claude/skills/session-continuity/protocols/04-pattern-extraction.md` and follow the **Pattern Extraction Protocol**. Reflect on what worked, what didn't, and log reusable patterns to `memory/patterns.md`. Skip only if the task was trivial (routine, nothing new learned).
+- **Session Close**: Read `.claude/skills/session-continuity/protocols/05-session-close.md` and follow the **Session Close Protocol**. Write a session log to `.claude/progress/sessions/` so the next session can resume cleanly.
 
 > These steps are **not optional**. They are what differentiate a pipeline that gets
 > smarter over time from one that repeats the same mistakes.
@@ -46,7 +46,7 @@ After completing a workflow or substantial task:
 ## 6. Recommend Next Step (MANDATORY)
 
 Before reporting completion via `notify_user`:
-- Check the pipeline progress state (`.agent/progress/spec-pipeline.md` or equivalent)
+- Check the pipeline progress state (`.claude/progress/spec-pipeline.md` or equivalent)
 - Propose the next valid pipeline command to the user
 - If multiple valid next steps exist, list them with brief context
 - Never end a conversation without telling the user what to do next

package/template/.claude/kit-sync.md

@@ -0,0 +1,6 @@
+# Kit Sync State
+
+upstream: https://github.com/RepairYourTech/cfsa-antigravity
+last_synced_commit: 515f87acc1239e276a85bf46b0d33e8925659a57
+last_synced_at: 2026-04-10T05:15:43Z
+kit_version: 2.19.4

package/template/.claude/memory/README.md

@@ -93,6 +93,36 @@ This memory system complements Claude Code's built-in features:
 - **Memory**: Decision/blocker/pattern tracking is CFSA-specific, stored here
 - **Sessions**: Session logs provide richer context than Claude Code's history
 
+## Native Memory Bridge (Required)
+
+When running in Claude Code, memory capture is dual-write:
+
+1. Write CFSA operational memory to `.claude/memory/*`
+2. Also sync durable collaboration memory to Claude native memory store using the native memory types
+
+### Mapping Rules
+
+- `patterns.md` entries map to native **feedback** memory when they reflect user corrections/preferences
+- `decisions.md` entries map to native **project** memory for project decisions; map to **user** memory when the decision is actually a user preference
+- `blockers.md` entries map to native **project** memory when they affect ongoing work context
+- External system pointers discovered while logging decisions/blockers map to native **reference** memory
+
+### Source of Truth
+
+- `.claude/memory/*` remains the pipeline execution log
+- Native memory remains the long-lived collaboration memory across sessions
+- Do not replace one with the other; keep them synchronized when trigger conditions are met
+
+### What Not To Sync
+
+- Trivial implementation details (file names, one-off variable choices)
+- Routine entries with no future value
+- Any codebase facts directly derivable from files or git history
+
+### Trigger
+
+On any memory-capture trigger (decision/correction/pattern/blocker), evaluate whether a native memory entry should also be written and sync it in the same conversation.
+
 ## Cleanup
 
 Session logs are automatically cleaned during template build, keeping only README.md. Decision, pattern, and blocker logs persist across installations.
@@ -101,4 +131,4 @@ Session logs are automatically cleaned during template build, keeping only READM
 
 - `.claude/skills/utilities/memory-protocol.md` — Memory management skill
 - `.claude/skills/utilities/progress-tracking.md` — Task integration
-- `.agent/progress/` — Antigravity's equivalent markdown-based tracking
+- `.claude/progress/` — Antigravity's equivalent markdown-based tracking

package/template/.claude/memory/blockers.md

@@ -0,0 +1,10 @@
+# Blockers
+
+Capture active and resolved blockers that affect progress.
+
+## Entry format
+
+### BLOCKER-001: [Description] (YYYY-MM-DD)
+- **Status**: active | resolved
+- **Impact**: What this blocks
+- **Resolution**: How it was resolved (if resolved)

package/template/.claude/memory/decisions.md

@@ -0,0 +1,19 @@
+# Decisions
+
+Capture non-trivial decisions with downstream impact.
+
+## Entry format
+
+### DEC-001: [Decision summary] (YYYY-MM-DD)
+- **Problem**: What needed deciding
+- **Options considered**: At least 2
+- **Decision**: What was chosen and why
+- **Downstream**: What this affects
+- **Reversibility**: High | Medium | Low
+
+### DEC-001: `.claude` owns its own runtime assets and state (2026-04-10)
+- **Problem**: `.claude` was coupled to `.agent` through symlinked skill-library content, `.agent/*` runtime references, and parity enforcement that made Claude non-standalone.
+- **Options considered**: Keep `.agent` as the canonical runtime backing for Claude; make `.claude` fully standalone while preserving build/check safeguards.
+- **Decision**: Make `.claude` self-contained. Runtime assets, progress state, and sync state now live under `.claude/`, while build/integrity scripts enforce standalone Claude invariants instead of `.agent` parity requirements.
+- **Downstream**: Affects `.claude` workflows, instructions, session continuity protocols, skill provisioning, template build behavior, and integrity verification.
+- **Reversibility**: Medium

package/template/.claude/memory/patterns.md

@@ -0,0 +1,26 @@
+# Patterns
+
+Capture repeatable best practices and anti-patterns discovered during execution.
+
+## Entry format
+
+### PAT-001: [Short description] (YYYY-MM-DD)
+- **Type**: best-practice | anti-pattern
+- **Confidence**: 0.5
+- **Context**: Where this applies
+- **Pattern**: What to do (or avoid)
+- **Source**: What triggered this entry
+
+### PAT-001: Make `.claude` standalone, not a `.agent` wrapper (2026-04-10)
+- **Type**: anti-pattern
+- **Confidence**: 0.5
+- **Context**: When porting or mirroring Antigravity functionality into the Claude runtime for this kit
+- **Pattern**: Do not treat `.claude` as a thin adapter over `.agent`. Claude assets, runtime state, and skill library ownership must live under `.claude/`, with parity/build checks handled as maintainer safeguards rather than runtime dependency.
+- **Source**: User correction after `.claude` was implemented as a wrapper that depended on `.agent`
+
+### PAT-002: Verify standalone runtime by build, integrity check, and `.agent` reference scan (2026-04-10)
+- **Type**: best-practice
+- **Confidence**: 0.5
+- **Context**: When changing Claude runtime structure or decoupling Claude assets from Antigravity assets
+- **Pattern**: Validate the standalone Claude runtime with `npm run build`, `./scripts/check-template-integrity.sh`, and a direct search for `.agent/` references under `.claude/`.
+- **Source**: This standalone `.claude` migration work

package/template/.claude/memory/sessions/README.md

@@ -0,0 +1,5 @@
+# Session Logs
+
+Store session-close notes here for continuity between runs.
+
+Recommended file naming: `YYYY-MM-DD.md`.

package/template/.claude/rules/completion-checklist.md

@@ -31,21 +31,21 @@ A unit of work is only DONE when:
 
 1. **The Code**: Implementation meets the contract and acceptance criteria.
 2. **The Tests**: All tests pass (validation step is green).
-3. **The Tracking**: Progress markdown files in `.agent/progress/` are updated.
+3. **The Tracking**: Progress markdown files in `.claude/progress/` are updated.
    - Acceptance criteria marked `[x]`
    - Slice/Task marked `[x]`
    - Phase progress fractions updated (e.g., `3/10` → `4/10`)
    - Overall progress fractions updated
 4. **The Locks**: All task claims (`[!]` flags and `files:` blocks) are removed.
-5. **The Memory**: All three memory files in `.agent/progress/memory/` are checked:
+5. **The Memory**: All three memory files in `.claude/memory/` are checked:
    - `patterns.md` — Any patterns observed (what worked, what failed) logged per Protocol 04. **User corrections are logged as anti-patterns.**
    - `decisions.md` — Any non-trivial decisions logged per Protocol 06
    - `blockers.md` — Any blockers hit logged with status and impact
    - See rule: `memory-capture` for triggers and format
-6. **The Session Log**: A session close log exists in `.agent/progress/sessions/`.
-   - Follow `.agent/skills/session-continuity/protocols/05-session-close.md` — write what was accomplished, deferred, and where the next session should start
+6. **The Session Log**: A session close log exists in `.claude/progress/sessions/`.
+   - Follow `.claude/skills/session-continuity/protocols/05-session-close.md` — write what was accomplished, deferred, and where the next session should start
 7. **The Next Step**: The user is told what comes next in the pipeline.
-   - Read the pipeline progress file (`.agent/progress/spec-pipeline.md` or equivalent)
+   - Read the pipeline progress file (`.claude/progress/spec-pipeline.md` or equivalent)
    - Propose the next valid pipeline command based on current state
    - If multiple options exist, present them as a numbered list with brief context
    - If the pipeline stage is complete, say so explicitly

package/template/.claude/rules/extensibility.md

@@ -30,7 +30,7 @@ trigger: always_on
 
 ## Naming Conventions
 
-Follow the project's established conventions in `.agent/instructions/patterns.md`. When no convention exists yet, use the stack's community standard.
+Follow the project's established conventions in `.claude/instructions/patterns.md`. When no convention exists yet, use the stack's community standard.
 
 ## Anti-Spaghetti Rules
 

package/template/.claude/rules/memory-capture.md

@@ -60,12 +60,25 @@ The pipeline has three memory files (`memory/patterns.md`, `memory/decisions.md`
 - Isolated implementation decisions (variable names, file paths) — skip
 - Repeating an existing pattern already logged — update confidence instead
 
+## Native Memory Sync (Claude Code)
+
+When running in Claude Code, memory capture is dual-write:
+
+1. Write the CFSA entry to `.claude/memory/{patterns|decisions|blockers}.md`
+2. Evaluate native-memory relevance and sync to Claude native memory store:
+   - User correction/preference → `feedback`
+   - User-specific preference/profile signal → `user`
+   - Project decision/blocker context with downstream impact → `project`
+   - External system pointer/where-to-look info → `reference`
+
+If a trigger requires CFSA logging and native-memory relevance, both writes are required in the same conversation.
+
 ## Pre-Completion Check
 
 Before calling `notify_user` to report completion of ANY workflow or substantial task:
 
 1. **Scan this conversation** for triggers in the table above
-2. **If triggers found** → write entries to the appropriate memory files
+2. **If triggers found** → write entries to `.claude/memory/*` and sync relevant items to native memory
 3. **If no triggers** → explicitly confirm: "No new patterns, decisions, or blockers to log"
 
 This check is **not skippable**. It applies to every pipeline stage, every conversation, every session.

package/template/.claude/rules/vertical-slices.md

@@ -41,7 +41,7 @@ When a multi-surface feature launches on one surface first:
 
 ## The Four Implementation Layers
 
-> **Note on terminology**: The layers below describe implementation completeness criteria — what every feature slice requires regardless of surface type. These are not the same as surface types (web/mobile/cli/etc.). For the mapping between surface types and how each implementation layer manifests on that surface, see `.agent/skills/prd-templates/references/surface-model.md`.
+> **Note on terminology**: The layers below describe implementation completeness criteria — what every feature slice requires regardless of surface type. These are not the same as surface types (web/mobile/cli/etc.). For the mapping between surface types and how each implementation layer manifests on that surface, see `.claude/skills/prd-templates/references/surface-model.md`.
 
 | Surface | What It Is | Example |
 |---------|-----------|---------|

package/template/.claude/skill-library/MANIFEST.md

@@ -8,7 +8,7 @@
 1. A pipeline skill (e.g., `/create-prd`, `/iterate-plan`, `/add-feature`) makes a tech decision
 2. It calls `/bootstrap-agents` with the new stack key=value
 3. Bootstrap reads this manifest and finds matching skills
-4. Matching skills are copied from `.agent/skill-library/` → `.agent/skills/`
+4. Matching skills are copied from `.claude/skill-library/` → `.claude/skills/`
 5. Any `{{PLACEHOLDER}}`s inside the copied skills are filled with current values
 
 ## Stack Triggers
@@ -98,10 +98,10 @@ When a stack key matches a value pattern (case-insensitive), install the listed
 | `UI_LIBRARY` | `*svelte-flow*` OR `*svelteflow*` OR `*xyflow*` | `stack/ui/svelte-flow` | `svelte-flow` |
 | `FRONTEND_FRAMEWORK` | `*react*` | `stack/ui/react-best-practices` | `react-best-practices` |
 | `FRONTEND_FRAMEWORK` | `*react*` | `stack/ui/react-composition-patterns` | `react-composition-patterns` |
-| `DESIGN_DIRECTION` | `*` (any confirmed value) | Fills `{{PLACEHOLDER}}`s in `.agent/skills/brand-guidelines/SKILL.md` in-place | `brand-guidelines` (in-place fill, not copy) |
+| `DESIGN_DIRECTION` | `*` (any confirmed value) | Fills `{{PLACEHOLDER}}`s in `.claude/skills/brand-guidelines/SKILL.md` in-place | `brand-guidelines` (in-place fill, not copy) |
 | `DESIGN_DIRECTION` | `*cinematic*` OR `*immersive*` | `stack/3d/threejs-pro` | `threejs-pro` |
 
-Note: `DESIGN_DIRECTION` does not copy a skill from the library — it fills placeholders in the pre-placed `.agent/skills/brand-guidelines/SKILL.md` template. Exception: when the direction is `Cinematic/Immersive`, `threejs-pro` is also installed as a standard skill copy, since WebGL and Three.js capabilities will be required.
+Note: `DESIGN_DIRECTION` does not copy a skill from the library — it fills placeholders in the pre-placed `.claude/skills/brand-guidelines/SKILL.md` template. Exception: when the direction is `Cinematic/Immersive`, `threejs-pro` is also installed as a standard skill copy, since WebGL and Three.js capabilities will be required.
 
 ### Authentication
 
@@ -168,7 +168,7 @@ Note: `DESIGN_DIRECTION` does not copy a skill from the library — it fills pla
 |-----------|--------------|-------------|-------------|
 | `OBSERVABILITY` | `*opentelemetry*` OR `*otel*` | `stack/observability/opentelemetry` | `opentelemetry` |
 | `OBSERVABILITY` | `*distributed-tracing*` OR `*jaeger*` OR `*zipkin*` | `stack/observability/distributed-tracing` | `distributed-tracing` |
-| `OBSERVABILITY` | `*structured-logging*` OR `*pino*` OR `*winston*` | Pre-installed: `.agent/skills/logging-best-practices` | `logging-best-practices` (pre-installed) |
+| `OBSERVABILITY` | `*structured-logging*` OR `*pino*` OR `*winston*` | Pre-installed: `.claude/skills/logging-best-practices` | `logging-best-practices` (pre-installed) |
 | `OBSERVABILITY` | `*python*` | `stack/observability/python-observability` | `python-observability` |
 | `OBSERVABILITY` | `*datadog*` | `stack/observability/datadog` | `datadog` |
 | `OBSERVABILITY` | `*prometheus*` OR `*grafana*` | `stack/observability/prometheus-grafana` | `prometheus-grafana` |
@@ -362,7 +362,7 @@ When the project includes a surface type, install the listed skills.
 | `api` | `surface/api/api-caching` | `api-caching` |
 | `api` | `surface/api/api-documentation-openapi` | `api-documentation-openapi` |
 | `api` | `surface/api/webhook-design` | `webhook-design` |
-| `api` | Pre-installed: `.agent/skills/api-design-principles` | `api-design-principles` (pre-installed) |
+| `api` | Pre-installed: `.claude/skills/api-design-principles` | `api-design-principles` (pre-installed) |
 | `api` | `surface/api/api-security-checklist` | `api-security-checklist` |
 | `api` | `stack/security/input-sanitization` | `input-sanitization` |
 
@@ -420,7 +420,7 @@ These skills are NOT auto-installed. Install via `/find-skills` or manually copy
 | `meta/mcp-builder` | Building MCP servers |
 | `meta/tmux-processes` | Long-lived process management via tmux |
 | `meta/using-tmux-for-interactive-commands` | Interactive CLI tools via tmux |
-| Pre-installed: `.agent/skills/brand-guidelines` | Brand color and typography application (pre-installed) |
+| Pre-installed: `.claude/skills/brand-guidelines` | Brand color and typography application (pre-installed) |
 | `meta/product-marketing-context` | Marketing context document generator — run first before any SEO or CRO skill |
 
 ---
@@ -441,7 +441,7 @@ These skills are specific to Claude Code's plugin and agent system. Not applicab
 
 To add a new skill to the library:
 
-1. Create a directory under the appropriate category: `.agent/skill-library/[category]/[subcategory]/[skill-name]/SKILL.md`
+1. Create a directory under the appropriate category: `.claude/skill-library/[category]/[subcategory]/[skill-name]/SKILL.md`
 2. Add a row to the appropriate trigger table above
 3. Use `{{PLACEHOLDER}}`s for any values that should be filled by bootstrap
 4. Test by running `/bootstrap-agents` with a matching stack value

package/template/.claude/skill-library/README.md

@@ -9,8 +9,8 @@
 
 1. A pipeline workflow (e.g., `/create-prd`) makes a tech stack decision
 2. It calls `/bootstrap-agents` with the new stack key=value
-3. Bootstrap reads `.agent/skill-library/MANIFEST.md` and finds matching skills
-4. Matching skills are copied from `.agent/skill-library/` → `.agent/skills/`
+3. Bootstrap reads `.claude/skill-library/MANIFEST.md` and finds matching skills
+4. Matching skills are copied from `.claude/skill-library/` → `.claude/skills/`
 5. Any `{{PLACEHOLDER}}`s inside the copied skills are filled with current values
 
 ## Installing Skills
@@ -26,7 +26,7 @@ Use `/find-skills` to search the library and install matching skills interactive
 
 ### Direct
 
-Copy a skill directory from `.agent/skill-library/[path]/` to `.agent/skills/[name]/`
+Copy a skill directory from `.claude/skill-library/[path]/` to `.claude/skills/[name]/`
 and manually replace any `{{PLACEHOLDER}}` values.
 
 ## Library Structure

package/template/.claude/skill-library/meta/claude-code/README.md

@@ -6,4 +6,4 @@ These skills are specific to Claude Code's plugin and agent system. They are not
 - **plugin-structure** — Claude Code plugin architecture (.claude-plugin/plugin.json)
 - **agent-development** — Claude Code agent frontmatter (model, color, triggering examples)
 
-These are manual-install only. Copy the relevant skill to `.agent/skills/` when working in Claude Code.
+These are manual-install only. Copy the relevant skill to `.claude/skills/` when working in Claude Code.

package/template/.claude/skill-library/surface/web/design-reference-data/SKILL.md

@@ -48,7 +48,7 @@ After the user confirms their design direction, run these searches to get concre
 #### 1. Find matching design styles
 
 ```bash
-python .agent/skills/design-reference-data/scripts/search.py \
+python .claude/skills/design-reference-data/scripts/search.py \
   "<direction + project context>" \
   --domain style --max-results 5
 ```
@@ -58,7 +58,7 @@ Example: `"modern dark mode SaaS dashboard"` or `"warm friendly wellness app"`
 #### 2. Find specific color palettes
 
 ```bash
-python .agent/skills/design-reference-data/scripts/search.py \
+python .claude/skills/design-reference-data/scripts/search.py \
   "<mood + industry>" \
   --domain color --max-results 3
 ```
@@ -68,7 +68,7 @@ Example: `"professional trust-building fintech"` or `"vibrant playful consumer"`
 #### 3. Find typography pairings
 
 ```bash
-python .agent/skills/design-reference-data/scripts/search.py \
+python .claude/skills/design-reference-data/scripts/search.py \
   "<style + readability needs>" \
   --domain typography --max-results 3
 ```
@@ -78,7 +78,7 @@ Example: `"clean serif editorial"` or `"modern sans-serif technical"`
 #### 4. Generate full design system recommendation
 
 ```bash
-python .agent/skills/design-reference-data/scripts/search.py \
+python .claude/skills/design-reference-data/scripts/search.py \
   "<project description>" \
   --design-system -p "ProjectName"
 ```

package/template/.claude/skill-library/surface/web/frontend-design/SKILL.md

@@ -14,7 +14,7 @@ This skill guides creation of **production-grade frontend interfaces that feel A
 
 ## Prerequisite: Read Brand Guidelines First
 
-Before making any aesthetic decisions, read `.agent/skills/brand-guidelines/SKILL.md`.
+Before making any aesthetic decisions, read `.claude/skills/brand-guidelines/SKILL.md`.
 
 - If the design direction is confirmed (no `{{PLACEHOLDER}}` values), **execute that direction**. Do not invent a new direction — the user confirmed this one during `/create-prd`.
 - If `{{DESIGN_DIRECTION}}` is still a literal placeholder, stop and tell the user: "Design direction hasn't been confirmed yet — run `/create-prd` first before building UI."

package/template/.claude/skill-library/surface/web/frontend-verification/SKILL.md

@@ -36,7 +36,7 @@ A 750-line Python scanner with 75+ automated checks across 11 categories:
 After implementing a frontend slice, run the audit on **changed files only**:
 
 ```bash
-python .agent/skills/frontend-verification/scripts/ux_audit.py src/components/NewFeature/ --json
+python .claude/skills/frontend-verification/scripts/ux_audit.py src/components/NewFeature/ --json
 ```
 
 **Exit codes:**
@@ -47,7 +47,7 @@ python .agent/skills/frontend-verification/scripts/ux_audit.py src/components/Ne
 For slice-level scans, you may skip categories unrelated to the current work:
 
 ```bash
-python .agent/skills/frontend-verification/scripts/ux_audit.py src/components/Chart/ \
+python .claude/skills/frontend-verification/scripts/ux_audit.py src/components/Chart/ \
   --skip-psychology --skip-typography
 ```
 
@@ -56,7 +56,7 @@ python .agent/skills/frontend-verification/scripts/ux_audit.py src/components/Ch
 At phase validation, run the audit on the **entire frontend source**:
 
 ```bash
-python .agent/skills/frontend-verification/scripts/ux_audit.py src/ --json
+python .claude/skills/frontend-verification/scripts/ux_audit.py src/ --json
 ```
 
 **Validation threshold:**

package/template/.claude/skills/bootstrap-agents/SKILL.md

@@ -9,11 +9,11 @@ version: 3.0.0
 **This is a utility command, not an entry point.** It gets called by other pipeline commands (like `/create-prd`, `/write-be-spec`, `/write-fe-spec`, `/implement-slice`) whenever they make tech stack decisions or introduce new dependencies.
 
 Bootstrap does two things:
-1. **Fill the surface stack map** in `.agent/instructions/tech-stack.md` with stack decisions
+1. **Fill the surface stack map** in `.claude/instructions/tech-stack.md` with stack decisions
 2. **Provision skills** from `skill-library/` using the 4-tier resolution chain
 
 **Input**: Surface-keyed stack values + optional global values
-**Output**: Filled map cells + newly installed skills in `.agent/skills/`
+**Output**: Filled map cells + newly installed skills in `.claude/skills/`
 
 ---
 
@@ -100,7 +100,7 @@ If any values are missing, leave those cells empty — they'll be filled on a fu
 
 ## 2. Fill the surface stack map
 
-Open `.agent/instructions/tech-stack.md` and update:
+Open `.claude/instructions/tech-stack.md` and update:
 
 ### 2a. Per-Surface Skills table
 
@@ -135,23 +135,23 @@ Replace in **both** `AGENTS.md` and `GEMINI.md`:
 
 ## 4. Fill instruction templates
 
-### `.agent/instructions/commands.md`
+### `.claude/instructions/commands.md`
 Write per-surface command sections. For each surface in the map, create a section with its commands (Test Cmd, Validation Cmd, Lint Cmd, Build Cmd, Dev Cmd, Package Mgr).
 
-### `.agent/instructions/workflow.md`
+### `.claude/instructions/workflow.md`
 Fill `{{VALIDATION_COMMAND}}` with the primary surface's validation command.
 
-### `.agent/instructions/patterns.md`
+### `.claude/instructions/patterns.md`
 Replace `{{FRAMEWORK_PATTERNS}}` if provided.
 
-### `.agent/instructions/structure.md`
+### `.claude/instructions/structure.md`
 Replace `{{PROJECT_STRUCTURE}}`, `{{ARCHITECTURE_TABLE}}` if provided.
 
 ---
 
 ## 5. Fill operational skill and rule templates
 
-Scan `.agent/skills/*/SKILL.md` and `.agent/rules/*.md` for `{{PLACEHOLDER}}` values and fill any that match the provided values. Currently applicable:
+Scan `.claude/skills/*/SKILL.md` and `.claude/rules/*.md` for `{{PLACEHOLDER}}` values and fill any that match the provided values. Currently applicable:
 
 - `{{VALIDATION_COMMAND}}` — in `fix-bug`, `main-workflow`, `deploy`, `refactor`
 - `{{PACKAGE_MANAGER}}` — in `refactor`, `security-audit`
@@ -173,13 +173,13 @@ If `skill-library/MANIFEST.md` does not exist, skip steps 7-8 and go to step 9.
 For each skill name referenced in ANY cell of the surface stack map, resolve it using this chain:
 
 ### Tier 1: Exact Match
-Check `.agent/skill-library/{name}/` (or `.agent/skills/{name}/` if already installed).
-- **Found in library** AND not yet installed → copy to `.agent/skills/{name}/`, fill any `{{PLACEHOLDER}}`s in the copied SKILL.md
+Check `.claude/skill-library/{name}/` (or `.claude/skills/{name}/` if already installed).
+- **Found in library** AND not yet installed → copy to `.claude/skills/{name}/`, fill any `{{PLACEHOLDER}}`s in the copied SKILL.md
 - **Already installed** → skip (idempotent)
 - **Not found** → proceed to Tier 2
 
 ### Tier 2: Partial Match + Adequacy Check
-Search `.agent/skill-library/` and `.agent/skills/` for skills whose name contains the base term. E.g., for `surrealdb-embedded`, check if `surrealdb` exists.
+Search `.claude/skill-library/` and `.claude/skills/` for skills whose name contains the base term. E.g., for `surrealdb-embedded`, check if `surrealdb` exists.
 
 - **Partial match found** → Read its `SKILL.md`. Assess: does it cover the needed variant? Check for keywords related to the specific need (e.g., "embedded", "WASM", "Rust-native" for `surrealdb-embedded`).
   - **Adequate** — the existing skill covers the variant → use it. Note in the report: `"{name}" resolved by existing "{partial}" skill (covers {variant})`.
@@ -187,7 +187,7 @@ Search `.agent/skill-library/` and `.agent/skills/` for skills whose name contai
 - **No partial match** → proceed to Tier 3
 
 ### Tier 3: External Discovery
-Read `.agent/skills/find-skills/SKILL.md` and invoke its discovery methodology to search for the skill externally.
+Read `.claude/skills/find-skills/SKILL.md` and invoke its discovery methodology to search for the skill externally.
 
 - **Found** → install → fill map cell with resolved name
 - **Not found** → proceed to Tier 4
@@ -252,7 +252,7 @@ Present the results to the calling command (not directly to the user — the cal
 Bootstrap is safe to call multiple times:
 
 - **Already-filled map cells**: Cells with values are NOT overwritten unless the calling command explicitly re-provides a value for that surface + column
-- **Already-installed skills**: Skills that already exist in `.agent/skills/` are not re-copied from the library
+- **Already-installed skills**: Skills that already exist in `.claude/skills/` are not re-copied from the library
 - **New surface rows**: New surfaces trigger new row creation without affecting existing rows
 - **Appending values**: Cross-cutting skills and accumulated columns append to existing comma-separated lists without duplicating
 - **Partial invocation**: Bootstrap can be called with just one surface or one key — it only fills what's provided