**Fundamental Principle**: *"Quality with focused context. Load only what is necessary, when it is necessary."*

---

## Purpose

`ac-lite` keeps OpenSpec rigor while reducing context load by:

1. Loading only mandatory core workflow skills by default.
2. Activating mobile/platform specialist skills only when objective gates trigger.
3. Enforcing hard implementation and closure checkpoints.

---

## Default Skill Set (Always Load)

1. `acfm-spec-workflow`
2. `acfm-memory`
3. `openspec-new-change` OR `openspec-ff-change`
4. `openspec-continue-change`
5. `openspec-apply-change`
6. `openspec-verify-change`
7. `openspec-archive-change`
8. `mobile-design`

---

## Persistent Memory Protocol (Always Active)

At session start:
1. Recall project memory before planning or implementation.
2. Search for task-specific patterns (architecture, bugfixes, conventions).
3. Treat recalled memory as active unless repo state or user instructions override.

Save reusable context after major decisions, fixes, and closures.

Fallback CLI examples:
```bash
acfm memory recall
acfm memory search "offline sync"
acfm memory stats
```

---


## SynapseGrid Collaboration Protocol (Optional but Recommended for Complex Tasks)

Use SynapseGrid when a task benefits from role-based collaboration and explicit review loops.

**Delegate to SynapseGrid when:**
- Scope spans planning + implementation + review
- Risk is medium/high (security, migrations, API contract changes)
- You need auditable outputs (transcript, meeting summary, artifacts)
- A single-pass implementation is likely to miss edge cases

**Role delegation map:**
- `planner` -> plan/constraints/acceptance criteria
- `critic` -> risk analysis and challenge assumptions
- `coder` -> implementation and concrete edits
- `reviewer` -> final verification and readiness checks

**Preferred MCP flow:**
1. `collab_start_session`
2. `collab_invoke_team`
3. `collab_wait_run`
4. `collab_get_result`
5. Optional: `collab_get_transcript`, `collab_get_meeting_log`, `collab_status`

**CLI fallback:**
```bash
acfm agents setup
acfm agents runtime set auto
acfm agents doctor --verbose
acfm agents start --task "..." --mux auto
acfm agents transcript --role all --limit 80
acfm agents summary
acfm agents artifacts --watch --interval 1200
```

**Model and runtime controls:**
```bash
acfm agents model list
acfm agents model choose
acfm agents runtime get
acfm agents runtime install-zellij
```

**Artifacts to inspect:**
- `~/.acfm/synapsegrid/<sessionId>/meeting-log.md`
- `~/.acfm/synapsegrid/<sessionId>/meeting-summary.md`
- `~/.acfm/synapsegrid/<sessionId>/turns/raw/*.ndjson`

**Lite rule:** delegate only when collaboration adds value; otherwise continue with normal `ac-lite` flow.


## Conditional Skill Gates

### Platform/Stack Gate
Load based on selected implementation stack:
- `flutter-expert` [IF Flutter]
- `ios-developer` [IF native iOS]
- `react-native-architecture` [IF React Native architecture/native modules]
- `react-native-design` [IF React Native UI/navigation/animation]
- `react-state-management` [IF global/complex state]

### Platform UI Gate
- `mobile-ios-design` [IF iOS UI changes]
- `mobile-android-design` [IF Android UI changes]

### Security Gate
- `vibe-security` [MANDATORY final audit before closure]

### Clarity/Scale Gate
- `spec-clarification` [IF requirements ambiguous]
- `spec-analysis` [IF cross-artifact consistency needs verification]
- `microtask-decomposition` [IF tasks remain too complex]
- `project-index` / `context-synthesizer` [IF repo/session scale requires]

### Research Gate
- `research-retrieval` [IF external docs are needed]

### Debug Gate
- `systematic-debugging` [IF blocked by bugs/regressions]

### Delivery Gate
- `documentation`, `sync-index`, `changelog-generator` [during closure]

---

## Mandatory Quality Gates

### Gate A: Ready to Implement
Before `openspec-apply-change`, all must be true:
- Change is initialized and valid (`acfm spec status --change <name> --json`)
- `tasks.md` exists with actionable checkboxes
- Platform and stack are explicit
- Required conditional skills were executed
- Session-start memory recall completed

### Gate B: Ready to Close
Before `openspec-archive-change`, all must be true:
- `openspec-verify-change` completed
- `vibe-security` executed as final security validation
- No unresolved CRITICAL/HIGH findings remain (or explicitly accepted with warning)
- Task list reflects final state
- Reusable context saved to memory
- `documentation` / `sync-index` / changelog steps executed as required

---

## Lite Workflow

### New Mobile Change
1. `acfm-spec-workflow` status/init checks.
2. Session-start memory recall (`acfm-memory`).
3. Start change (`openspec-new-change` or `openspec-ff-change`).
4. Build artifacts (`openspec-continue-change`).
5. Trigger and run only needed conditional skills.
6. Pass Gate A.
7. Implement (`openspec-apply-change`).
8. Verify (`openspec-verify-change`).
9. Run `vibe-security` for final security audit and remediate findings.
10. Pass Gate B.
11. Archive/sync and generate closure docs.

### Existing Mobile Change
1. Confirm initialization and target change.
2. Run memory recall.
3. Refresh artifacts.
4. Run only triggered conditional skills.
5. Pass Gate A.
6. Implement and debug as needed.
7. Verify.
8. Run `vibe-security` for final security audit and remediate findings.
9. Pass Gate B.
10. Archive/sync.

---

## Operational Rules

1. Do not load non-triggered skills "just in case".
2. Do not implement before platform/stack is explicit.
3. If risk increases, load matching skill immediately.
4. Keep outputs concise and auditable (which gates passed/failed and why).
5. Save reusable memory at significant milestones.