cfsa-antigravity 2.7.0 → 2.8.0

package/template/.agent/workflows/ideate-extract.md

@@ -19,181 +19,85 @@ pipeline:
 
 Classify the user's input, create the fractal `ideation/` folder structure, and load skills.
 
-**Prerequisite**: If standalone, verify user provided input via `@file` or verbal. If none, prompt: \"What would you like to build? Provide a file or describe your idea.\"
+**Prerequisite**: If standalone, verify user provided input via `@file` or verbal. If none, prompt: "What would you like to build? Provide a file or describe your idea."
+
+If `@file` was provided:
+- **If file does not exist or is not readable** → **STOP**: "File not found at `[path]`. Provide a valid file path or describe your idea verbally."
+- **If file is empty** (0 bytes) → **STOP**: "File at `[path]` is empty. Provide a file with content or describe your idea verbally."
+- **If file is readable** → proceed to Step 1.
 
 ---
 
 ## 1. Input assessment
 
-Classify what the user has provided. This determines which mode the `idea-extraction` skill uses.
-
-| Input Type | Detection | Extraction Mode |
-|---|---|---|
-| **Rich document** | >5KB, detailed docs, design conversations, prior specs | Extraction |
-| **Thin document** | <5KB, structured but shallow (bullet list, rough PRD) | Expansion |
-| **Conversational dump** | Chat logs, unstructured conversation transcripts | Extraction (with noise filtering) |
-| **Verbal / one-liner** | User describes idea in chat, no files | Interview |
-| **Nothing** | "I want to build an app" or similar | Interview (deep) |
-
-**For rich inputs (Extraction mode):**
-1. Read/ingest all provided documents
-2. **Proportionality check**: If source >50KB, total ideation output must be ≥30% of source line count
-3. **Interview the document**: Run the interview question framework from `idea-extraction/SKILL.md` Extraction Mode Phase 1. Extract concepts (not headings) with citations. Classify every concept through the Node Classification Gate. Build the classification table and proposed domain map.
-4. **BLOCKING GATE — Do NOT proceed to Step 1.5 until the classification table is built.** The classification table is a required artifact. No domain folders can be created without it.
-5. Identify gaps — questions the document doesn't answer → these become Phase 2 user interview questions
-6. Use idea-extraction skill to refine, not re-derive
-
-**For thin inputs (Expansion mode):**
-1. Read input and identify domain boundaries
-2. Identify depth level per domain (surface → detailed → implementation-ready)
-3. Note shallowest domains for priority treatment in `/ideate-discover`
-
-**For verbal / no input (Interview mode):**
-1. Read `.agent/skills/idea-extraction/SKILL.md` and enter Interview mode
-2. Start with: "In one sentence, what problem does this solve and for whom?"
-3. **Immediately after**: Run Structural Classification (Step 1.3)
-4. From that sentence + classification, identify initial domains
-5. Proceed to folder seeding (Step 1.5)
+Read `.agent/skills/prd-templates/references/input-classification.md` — classify the input into one of the five types.
 
-## 1.3. Structural Classification
+Read `.agent/skills/idea-extraction/SKILL.md` → `## Input-Adaptive Modes` — follow the process for the classified input type.
 
-Read the **Structural Classification Protocol** in `.agent/skills/idea-extraction/SKILL.md`.
+**BLOCKING GATE**: Classification must produce a single input type. If ambiguous between Rich and Thin, use the 5KB threshold. If ambiguous between Thin and Verbal, check whether a file was provided.
 
-This step determines the folder layout. It MUST run **before** any domain folders are created (Step 1.5).
+**For Extraction mode (rich input)**: Follow `idea-extraction/SKILL.md` → Extraction Mode Phase 1 steps 1-6.
 
-**For rich/thin document input:**
-1. Scan for surface signals (see detection table in skill)
-2. If multi-product signals detected, scan for hub-and-spoke vs peer signals:
-   - One surface described as "the platform" or "the API" → **hub-and-spoke**
-   - All surfaces access central database through one API → **hub-and-spoke**
-   - Surfaces described as equally independent → **peer**
-3. If no multi-product signals → **single-surface**
-4. If ambiguous → ask the user the classification questions
+**BLOCKING GATE — Do NOT proceed to Step 1.3 until the classification table is built.** The classification table (from skill Phase 1 step 4) is a required artifact. No domain folders can be created without it.
 
-**For interview / verbal input:**
-1. After the opening problem statement, ask:
-   - "Who are the distinct user types or audiences?"
-   - "What platforms does this need to live on?" (web, mobile, desktop, API, CLI)
-2. If multi-product → ask: "Is there a primary platform that others depend on, or are all surfaces independent peers?"
-3. Classify the project shape
+**For Expansion mode (thin input)**: Follow `idea-extraction/SKILL.md` → Expansion Mode steps 1-3. Note shallowest domains for priority treatment in `/ideate-discover`.
 
-**Record the classification** in `ideation-index.md` under `## Structural Classification` per the super-index template.
+**For Interview mode (verbal / no input)**: Follow `idea-extraction/SKILL.md` → Interview Mode steps 1-2. Proceed to Step 1.3 immediately after opening question.
+
+## 1.3. Structural Classification
 
-Four shapes: `single-surface`, `multi-surface-shared`, `multi-product-hub`, `multi-product-peer`.
+Read `.agent/skills/idea-extraction/SKILL.md` → `## Structural Classification Protocol`. Follow the protocol for the current input type (document vs interview).
+
+**BLOCKING GATE**: Classification must run BEFORE any domain folders are created (Step 1.5). Record the result (`single-surface`, `multi-surface-shared`, `multi-product-hub`, `multi-product-peer`) in `ideation-index.md` under `## Structural Classification`.
 
 ## 1.4. Re-run check
 
-Before seeding, check whether `docs/plans/ideation/ideation-index.md` already exists.
+Check whether `docs/plans/ideation/ideation-index.md` already exists.
 
-- If it **exists**: Present summary (expansion mode, domain count, depth markers). Ask: "An ideation folder already exists. **Continue** or **start fresh**?"
+- **Exists** → Present summary (expansion mode, domain count, depth markers). Ask: "An ideation folder already exists. **Continue** or **start fresh**?"
   - **Continue** → skip seeding, jump to Step 1.6
   - **Start fresh** → archive to `docs/plans/ideation-archive-[timestamp]/`, then seed
-- If it **does not exist**: proceed with seeding.
+- **Does not exist** → proceed with seeding.
 
 ## 1.4.5. Classification Confirmation (Extraction mode only)
 
 > **BLOCKING GATE**: Before creating ANY domain folder, you MUST complete this step.
-> This step exists because the test revealed that the Node Classification Gate was being
-> bypassed — the agent mirrored source document headings as domains instead of classifying
-> each concept. This gate makes classification a hard prerequisite for folder creation.
-
-1. **Verify the classification table exists.** Step 1 item 3 must have produced a completed classification table. If it doesn't exist, STOP — go back and run Extraction Mode Phase 1 from `idea-extraction/SKILL.md`.
-2. **Present the classification table and proposed domain map to the user.** Show:
-   - Every concept with its source location, gate result, and reasoning
-   - The proposed domain hierarchy (what folders would be created and where)
-   - The gap list (questions the document didn't answer)
-3. **Ask**: "Here's how I classified your content. Does this look right?"
-4. **Wait for user confirmation or corrections.** Do NOT proceed until the user responds.
-5. **Apply any corrections** the user provides — update the classification table and domain map.
-
-**The classification table must show for each concept:**
-- Concept name and source location (line numbers or section reference)
-- Gate result: domain / sub-domain / feature / cross-cut / not-a-product-domain
-- Gate reasoning: which gate questions were asked and how they were answered
-
-**Do NOT mirror source document headings as domains.** Source headings are hints for finding concepts. The Node Classification Gate determines what each concept is.
+
+1. **Verify the classification table exists.** If it doesn't exist → STOP — go back and run Extraction Mode Phase 1 from `idea-extraction/SKILL.md`.
+2. **Present the classification table and proposed domain map to the user.** Follow `idea-extraction/SKILL.md` → Extraction Mode Phase 2 step 7 for what to show.
+3. **Wait for user confirmation or corrections.** Do NOT proceed until the user responds.
+4. **Apply any corrections** the user provides.
+5. **If user rejects the classification entirely** (e.g., "these domains are completely wrong") → return to Step 1 and re-classify from scratch, incorporating the user's feedback as explicit constraints. Do NOT attempt to patch the existing classification — start the classification protocol over.
 
 ## 1.5. Seed fractal `ideation/` folder
 
-Read these templates:
-- `.agent/skills/prd-templates/references/ideation-index-template.md` (super-index)
-- `.agent/skills/prd-templates/references/ideation-crosscut-template.md` (global CX)
-- `.agent/skills/prd-templates/references/fractal-node-index-template.md` (node index)
-- `.agent/skills/prd-templates/references/fractal-cx-template.md` (node CX)
-- `.agent/skills/prd-templates/references/fractal-feature-template.md` (feature file)
+Read `.agent/skills/prd-templates/references/folder-seeding-protocol.md` and follow it.
 
 Read `.agent/skills/technical-writer/SKILL.md` and follow its methodology.
 
-**ADDITIVE ONLY** — The `docs/plans/ideation/` directory already exists in the kit with `.gitkeep` and `README.md`. You are ADDING files into this existing directory. Do NOT delete, overwrite, or replace the directory itself. Do NOT remove any existing files. Create new files alongside what already exists.
-
-**Create the base structure** (all project shapes). This is what the folder should contain AFTER seeding — kit-shipped files plus new pipeline files:
+**BLOCKING GATE**: If Step 1.4.5 was not completed (Extraction mode), STOP — you cannot seed without a confirmed classification.
 
-```
-docs/plans/ideation/
-├── .gitkeep                ← KIT-SHIPPED — do not touch
-├── README.md               ← KIT-SHIPPED — do not touch
-├── ideation-index.md       ← NEW: super-index (from ideation-index-template)
-├── ideation-cx.md          ← NEW: global cross-cuts (from ideation-crosscut-template)
-└── meta/                   ← NEW: created by this step
-    ├── problem-statement.md
-    ├── personas.md
-    ├── competitive-landscape.md
-    └── constraints.md
-```
+**Partial failure recovery**: If folder seeding fails mid-way (some domain folders created, others not):
+1. List what was successfully created vs. what failed
+2. Present to user: "Folder seeding partially failed. [N] of [M] domains created. Retry the remaining [M-N] domains?"
+3. If user says retry → seed only the missing domains
+4. If user says abort → delete all newly created folders and STOP
 
-**For multi-product projects**, additionally create `surfaces/` with sub-folders per surface.
-
-**Post-seeding verification gate**: After creating all files, verify that `docs/plans/ideation/.gitkeep` and `docs/plans/ideation/README.md` still exist. If EITHER file is missing → **STOP**: "Kit-shipped files were destroyed during seeding. Restore `.gitkeep` and/or `README.md` to `docs/plans/ideation/` before continuing."
-
-**Seed domains from CONFIRMED classification table** (not from source headings):
-
-> **CRITICAL**: The domain list comes from the classification table confirmed in Step 1.4.5.
-> Do NOT re-derive domains from the source document. Do NOT fall back to source headings.
-> If Step 1.4.5 was not completed, STOP — you cannot seed without a confirmed classification.
-
-For each concept classified as **domain** in the confirmed table:
-1. Determine placement from the structural classification:
-   - Single-surface → create `docs/plans/ideation/{NN}-{slug}/`
-   - Hub-and-spoke → surface-exclusive in `surfaces/{surface}/{NN}-{slug}/`, shared in hub surface
-   - Peer → surface-exclusive in `surfaces/{surface}/{NN}-{slug}/`, shared in `shared/{NN}-{slug}/`
-2. Create the domain folder with: `{slug}-index.md` + `{slug}-cx.md`
-3. For concepts classified as **sub-domain**, create sub-domain folders nested inside their parent domain
-4. For concepts classified as **feature**, create feature files inside their parent domain or sub-domain
-5. For concepts classified as **cross-cut**, add entries to the appropriate CX files (domain-level or global)
-6. For concepts classified as **not-a-product-domain**, add notes to `meta/constraints.md` for `/create-prd`
-7. Update `ideation-index.md` structure map with paths
-
-**Seeding content by input type:**
-
-- **Rich document**: Seed each domain/sub-domain/feature with content from the source document using the source citations in the classification table. Run fidelity check: every major concept in the source must map to SOMETHING in the output (domain, sub-domain, feature, CX entry, or `/create-prd` note). Add `> Source: path/to/original.md` to index.
-- **Chat transcript**: Noise filter → extract signal → seed domain folders with structured output.
-- **Thin document**: Create domain folders with depth markers on feature files.
-- **Verbal / one-liner**: Create domain folders with scaffolding. Feature files are `[SURFACE]`.
+**Post-seeding verification gate**: Verify `docs/plans/ideation/.gitkeep` and `docs/plans/ideation/README.md` still exist. If EITHER is missing → **STOP**: "Kit-shipped files were destroyed during seeding. Restore them before continuing."
 
 ## 1.6. Engagement Tier Selection (ALL input types)
 
-All three tiers are available for all input types. Present with the recommended default for this input type (Rich/Thin/Chat → Hybrid, Verbal → Interactive):
+Read `.agent/skills/prd-templates/references/engagement-tier-protocol.md` for tier definitions and behavior.
 
-> **How involved do you want to be?**
-> 1. 🤖 **Auto** — I explore independently via Deep Think. You review at the end.
-> 2. 🤝 **Hybrid** *(recommended)* — Structural stuff auto, product decisions pause for you.
-> 3. 💬 **Interactive** — I pause at every gate. Full interview mode.
+Present with the recommended default for this input type (Rich/Thin/Chat → Hybrid, Verbal → Interactive).
 
 **Wait for user answer.** Write the engagement tier to `ideation-index.md` under `## Engagement Tier` immediately.
 
 ## 1.6.5. Expansion Mode Selection (ALL input types)
 
-Present expansion mode options:
-
-1. **Full exploration** *(recommended for 3+ domains)* — Breadth-before-depth with Deep Think
-2. **Process as-is** — Proceed with what's captured
-3. **Expand vertically** — Drill deeper into existing features
-4. **Expand horizontally** — Add new domains
-5. **Cross-cutting concerns** — Map feature interactions
-6. **Combination** — User specifies dimensions and order
-7. **Audit ambiguity first** — Inline check before deciding
+Read `.agent/skills/prd-templates/references/expansion-modes.md` for mode options and default recommendations.
 
-**Wait for user answer.** Write expansion mode to `ideation-index.md` under `## Expansion Mode` immediately.
+Present the options. **Wait for user answer.** Write expansion mode to `ideation-index.md` under `## Expansion Mode` immediately.
 
 ## 2. Load skills
 
@@ -201,8 +105,8 @@ Read `.agent/skills/idea-extraction/SKILL.md` and follow its methodology through
 
 Also read `.agent/skills/resolve-ambiguity/SKILL.md` — use reactively when encountering ambiguity that can be resolved without user input.
 
-### Propose next step
+### Next step
 
-Proceed to `/ideate-discover` to explore domains using the recursive breadth-before-depth model with the fractal structure.
+**STOP** — do NOT proceed to any other workflow. The only valid next step is `/ideate-discover`.
 
-> If invoked standalone, surface this via `notify_user`. If invoked by parent `/ideate`, this is a natural handoff.
+> If invoked standalone, surface this via `notify_user` and wait for user confirmation before running `/ideate-discover`.

package/template/.agent/workflows/ideate-validate.md

@@ -19,154 +19,98 @@ pipeline:
 
 Explore constraints, verify domain exhaustion, and compile the vision summary.
 
-**Prerequisite**: If invoked standalone, verify `docs/plans/ideation/ideation-index.md` exists with leaf nodes at `[DEEP]` or `[EXHAUSTED]` level. If not, prompt the user to run `/ideate-discover` first.
+**Prerequisite**: If invoked standalone, verify `docs/plans/ideation/ideation-index.md` exists with leaf nodes at `[DEEP]` or `[EXHAUSTED]` level. If not → **STOP**: "Run `/ideate-discover` first."
 
 ---
 
 ## 7.5. Read Engagement Tier
 
-Read `## Engagement Tier` from `docs/plans/ideation/ideation-index.md`. Apply to gates in this shard:
+Read `## Engagement Tier` from `docs/plans/ideation/ideation-index.md`.
 
-- **Auto**: Steps 8 (constraints, metrics, competitive positioning) → agent self-interviews using Deep Think, writes answers to files. Step 10.5 (below) provides a review checkpoint before compilation.
-- **Hybrid**: Steps 8 product decisions → pause for user. Structural checks → auto.
-- **Interactive**: All steps pause for user confirmation.
+Read `.agent/skills/prd-templates/references/engagement-tier-protocol.md` — apply the tier's gate behavior for this shard.
 
 ## 8. Constraints and metrics
 
-Read `.agent/skills/idea-extraction/SKILL.md` and follow its Deep Think Protocol.
+Read `.agent/skills/idea-extraction/SKILL.md` → Deep Think Protocol.
 
-Read `.agent/skills/prd-templates/references/ideation-meta-template.md` for the constraints template.
+Read `.agent/skills/prd-templates/references/constraint-exploration.md` — follow the constraint questions, tier-specific behavior, success metrics, and competitive positioning procedures.
 
-Explore constraints with the user. Write to `docs/plans/ideation/meta/constraints.md`:
+**Showstopper detection**: If any constraint is classified as a fundamental viability blocker (e.g., regulatory impossibility, technical impossibility with current tech, market size < viable threshold) → **STOP**. Present to the user:
 
-1. **Budget** — Self-funded? VC-backed? Monthly infrastructure ceiling?
-2. **Timeline** — Launch target? Phased rollout?
-3. **Team** — Solo dev? Small team? Skill gaps?
-4. **Compliance** — GDPR, PCI, COPPA, HIPAA, SOC 2? Age restrictions?
-5. **Performance** — Expected scale (users, requests, data)? Latency requirements?
-6. **Surface classification validation** — Verify the structural classification from `ideation-index.md` (set in `ideate-extract` Step 1.3) still holds. Have any new surfaces been discovered during exploration? Has the project shape changed (e.g., what started as single-surface now has a mobile app too)? If the classification needs updating, update it now and note any domain files that need to be relocated.
+> ⚠️ **Potential showstopper identified**: [constraint description]
+>
+> Options:
+> 1. **Pivot** — modify the idea to avoid this constraint
+> 2. **Accept risk** — proceed knowing this constraint exists
+> 3. **Abandon** — this idea is not viable
 
-**Deep Think**: "Based on the product type and user personas, what constraints would I expect that haven't been mentioned? For example, does this product handle payments (PCI)? Does it serve minors (COPPA)? Does it store health data (HIPAA)?"
+Wait for user decision. If pivot → update problem statement and loop back to re-explore affected domains. If accept risk → document in `meta/constraints.md` as accepted risk and continue. If abandon → end the workflow.
 
-**Interactive/Hybrid**: Present each constraint question to user, wait for answers. Write each confirmed constraint to `meta/constraints.md` immediately.
-
-**Auto**: Self-interview using Deep Think. Write all answers with reasoning to `meta/constraints.md` immediately. Mark each answer as `[AUTO-CONFIRMED]` for traceability.
-
-If the surface classification changed, update `ideation-index.md` `## Structural Classification` section.
-
-### Success metrics
-
-For each persona, define concrete success metrics. Write to `ideation-index.md` (or link to domain files where the metric applies):
-
-- What metric proves this product solves the persona's problem?
-- What's the target number? (specific — not "good response times")
-- What's the measurement method?
-
-### Competitive positioning
-
-If not already explored in `/ideate-discover` Step 4, explore competitive landscape now. Write to `docs/plans/ideation/meta/competitive-landscape.md`:
-
-- Name 2-4 direct competitors
-- For each: what they do well, where they fail, how we differentiate
-- What's the moat? (network effects, data, expertise, switching costs)
+If the surface classification changed during constraint exploration, update `ideation-index.md` `## Structural Classification` section.
 
 ---
 
 ## 9. Domain exhaustion check
 
-This is the final validation gate before compilation.
-
-### Read the fractal tree
+Read `.agent/skills/prd-templates/references/domain-exhaustion-criteria.md` — apply all criteria and follow the execution procedure.
 
-Read `docs/plans/ideation/ideation-index.md` and recursively review:
-- Every node's status marker (surface → domain → sub-domain → feature)
-- All leaf feature files' status markers
-- All CX files at every level for pending entries
+If any criterion fails → take the specified action. If proportionality fails → return to `/ideate-discover` for under-explored areas.
 
-### Exhaustion criteria
+### 9.5. Domain Gap Reasoning (missing domain detection)
 
-| Check | Criteria | Action if Fail |
-|-------|----------|----------------|
-| All leaf nodes ≥ `[DEEP]` | Every feature file in the tree is `[DEEP]` or `[EXHAUSTED]` | Drill remaining feature files |
-| Status propagation correct | Parent nodes reflect their children's status | Update parent indexes |
-| All Must Have features ≥ Level 2 | Every Must Have has sub-features AND edge cases AND Role Lens | Deep Think + drill |
-| Deep Think zero hypotheses | Final Deep Think pass across ALL leaf nodes yields no new hypotheses | Present any new hypotheses, drill if confirmed |
-| All CX files clean | No Medium/Low confidence entries remain at any level — all are High or rejected | Run synthesis questions on pending pairs |
-| Role Lens complete | Every feature file has a populated Role Lens table | Fill missing Role Lens entries |
-| User confirmation | User explicitly confirms "nothing else" for each domain | Ask for each under-explored domain |
+After verifying existing domains are deep enough, reason about whether **entire domains are missing**:
 
-### Execute exhaustion check
+1. **Product archetype analysis**: Identify the product archetype (e.g., marketplace, SaaS tool, social platform, developer tool). List the standard domain categories for this archetype.
+2. **Gap identification**: Compare the standard domain list against the actual domain folders in `docs/plans/ideation/domains/`. List any standard domains with no corresponding folder.
+3. **Cross-feature gap detection**: Read all CX files. For each unresolved cross-cut, ask: "Does this cross-cut imply a domain that doesn't exist yet?" (Example: if multiple features reference "notifications" but no notifications domain exists, that's a structural gap.)
+4. **Present missing domains**:
 
-1. Walk the fractal tree. For each leaf node below `[DEEP]`:
-   - "Feature [X] in [domain] is still at [status]. Drill deeper or intentionally minimal?"
-   - If "drill" → return to `/ideate-discover`
-   - If "intentionally minimal" → note in feature file and proceed
+> 🏗️ **Potential missing domains:**
+>
+> Given this is a [product archetype], these domains are standard but not present:
+> - **[Domain A]** — [why it's expected: most [archetype] apps have this because...]
+> - **[Domain B]** — [why it's expected: cross-cuts between features X and Y imply this]
+>
+> These are suggestions. You may have intentionally excluded them. Want to add any?
 
-2. Run **final Deep Think pass**: For each `[DEEP]` leaf node, apply the four Deep Think questions. Present any new hypotheses.
-   - If confirmed → drill, update feature files
-   - If zero hypotheses → mark `[EXHAUSTED]`, propagate status upward
+**STOP** — wait for user response. For each accepted domain:
+1. Create the domain folder with index and CX files
+2. Run a quick Level 1 breadth sweep (from idea-extraction skill)
+3. Re-run the exhaustion check on the expanded domain set
 
-3. Walk ALL CX files at every level. Resolve any Medium/Low confidence entries.
+For rejected domains: add to `ideation-index.md` `## Considered & Rejected` section.
 
-4. Verify all feature files have populated Role Lens tables.
-
-5. Update `ideation-index.md` progress summary with final counts (total leaf nodes, exhausted count, CX entries confirmed).
-
----
-
-## 10. Vision deepening (if needed)
-
-After the exhaustion check, verify proportionality:
-
-- **Rich inputs**: Total domain file content (all files combined) should be at least 30% of the original source document's line count. If short, identify what was lost.
-- **All inputs**: Each domain with `[DEEP]` or `[EXHAUSTED]` status should have at least 3 sub-areas drilled with edge cases.
-
-If proportionality fails, return to `/ideate-discover` for the under-explored areas.
+**Loop guard**: Track how many times this shard has returned to `/ideate-discover` for exhaustion remediation.
+- **1st return** → normal. Run discover again on the flagged areas.
+- **2nd return** → warn: "This is the second remediation loop. Remaining gaps: [list]. Resolve these specifically or they will be escalated."
+- **3rd return** → **STOP**: "Exhaustion check has failed 3 times. Remaining gaps: [list]. Present these to the user as known gaps and ask: accept as-is, or manually provide the missing content?"
 
 ---
 
 ## 10.5. Auto Tier Review Checkpoint (Auto tier only)
 
-If engagement tier is **Auto**, present a comprehensive review before compilation:
+If engagement tier is **Auto**:
 
-1. **List all auto-confirmed decisions** with their Deep Think reasoning — from domain classification, feature drilling, constraints, personas, competitive positioning
-2. **Highlight any `[AUTO-CONFIRMED]` entries** in `meta/constraints.md`, `meta/personas.md`, and `meta/competitive-landscape.md`
-3. **Present for review**: "I explored your idea independently. Here's everything I decided and why. Override anything before I compile the vision."
-4. **Wait for user response.** Apply any overrides. Write corrections to files immediately.
+1. List all auto-confirmed decisions with their Deep Think reasoning
+2. Highlight any `[AUTO-CONFIRMED]` entries in `meta/constraints.md`, `meta/personas.md`, `meta/competitive-landscape.md`
+3. Present: "I explored your idea independently. Here's everything I decided and why. Override anything before I compile the vision."
+4. **Wait for user response.** Apply any overrides. Write corrections immediately.
 
-For **Hybrid** and **Interactive** tiers, skip this step — the user already confirmed during exploration.
+For **Hybrid** and **Interactive** tiers → skip this step.
 
 ---
 
 ## 11. Compile vision document
 
-Read `.agent/skills/prd-templates/references/vision-template.md` for the output template.
+Read `.agent/skills/prd-templates/references/vision-template.md` for the output template and required sections.
 
 Read `.agent/skills/technical-writer/SKILL.md` and follow its methodology.
 
-Compile `docs/plans/vision.md` as a **human-readable executive summary** of the ideation output. This is the "sales pitch" — it is NOT consumed by the pipeline. The pipeline reads `ideation-index.md` directly.
-
-**Vision.md contents (Option B — Executive Summary):**
-
-1. **Problem Statement** — Condensed from `meta/problem-statement.md`
-2. **Target Users** — Condensed persona summaries from `meta/personas.md` (name + role + pain point + success criteria — not the full 6-field exploration)
-3. **Solution Overview** — 2-3 paragraphs describing what the product does
-4. **Domain Map** — One paragraph per domain (condensed from domain files, not the full exploration)
-5. **MoSCoW Feature Matrix** — Feature names + domain links (not the drill-down details)
-6. **Key Differentiators + Competitive Landscape** — From `meta/competitive-landscape.md`
-7. **Constraints Summary** — From `meta/constraints.md`
-8. **Key Decisions** — Numbered list from `ideation-index.md` decision log
-
-Add a header note:
-
-```markdown
-> **This is a human-readable project summary.** For pipeline-grade detail, see
-> [ideation-index.md](ideation/ideation-index.md) and the domain files it references.
-```
+Compile `docs/plans/vision.md` as a human-readable executive summary. This is NOT consumed by the pipeline — the pipeline reads `ideation-index.md` directly.
 
 ### Fidelity check
 
-Verify that every domain in `ideation-index.md` appears in `vision.md`. Nothing dropped during compilation. This is a summary, not a filter.
+Verify every domain in `ideation-index.md` appears in `vision.md`. Nothing dropped during compilation.
 
 ---
 
@@ -174,30 +118,12 @@ Verify that every domain in `ideation-index.md` appears in `vision.md`. Nothing
 
 ### Self-check against Ideation rubric
 
-Before presenting to the user, self-check the ideation output:
-
-Read `.agent/skills/pipeline-rubrics/references/ideation-rubric.md` before applying the self-check dimensions.
+Read `.agent/skills/pipeline-rubrics/references/ideation-rubric.md` and apply all 12 dimensions as the self-check.
 
-| # | Dimension | Check |
-|---|-----------|-------|
-| 1 | Problem Clarity | Is the problem one sentence, specific, and testable? |
-| 2 | Persona Specificity | Are personas named with all 6 fields? |
-| 3 | Feature Completeness | Is MoSCoW complete? Are Must Haves at ≥Level 2 depth? |
-| 4 | Constraint Explicitness | Are all axes (budget, timeline, team, compliance, performance) addressed? |
-| 5 | Success Measurability | Are there concrete numbers/thresholds? |
-| 6 | Competitive Positioning | Are competitors named with differentiation? |
-| 7 | Open Question Resolution | Do all open questions have owners + deadlines? |
-| 8 | **Input-Output Proportionality** | Is the ideation output proportional to input richness? |
-| 9 | **Domain Coverage** | Are all domains at `[DEEP]` or `[EXHAUSTED]`? |
-| 10 | **Deep Think Coverage** | Were hypotheses tracked? Are all resolved (confirmed/rejected)? |
-| 11 | **Cross-Cut Completeness** | Is the ledger clean? No pending entries? |
-| 12 | **Fractal Structure Compliance** | Does every folder have an index + CX file? Do leaf nodes use the feature template? Does hub-and-spoke placement match classification? Are Role Matrix and Role Lens populated? |
+For any dimension that scores ⚠️ or ❌ → resolve it NOW. Loop back to the relevant step. Do not present a document with known gaps.
 
-For any dimension that scores ⚠️ or ❌, resolve it NOW — don't present a document with known gaps. Loop back to the relevant step and work through it with the user.
-
-> **Note**: This is an internal self-check, not a formal audit. For a rigorous,
-> independent audit with evidence citations, run `/audit-ambiguity ideation` as a
-> separate step after this workflow completes.
+**Remediation loop guard**: Track remediation attempts per dimension.
+- After **3 failed attempts** on the same dimension → **STOP**: "Dimension '[name]' has failed remediation 3 times. Presenting to user as a known gap with context: [what was tried, why it failed]." Include it in the review presentation as an unresolved item for user decision.
 
 ### Present for review
 
@@ -205,14 +131,12 @@ Use `notify_user` to request review of:
 - `docs/plans/ideation/ideation-index.md` — the pipeline key file
 - `docs/plans/vision.md` — the human summary
 
-Include:
-- Summary of the self-check results (all 12 dimensions)
-- Any areas where you resolved gaps during the self-check
-- The final domain coverage map
-- Count of Deep Think hypotheses: N presented, N confirmed, N rejected
+Include: self-check results (all 12 dimensions), any resolved gaps, final domain coverage map, Deep Think hypothesis counts.
+
+**STOP** — do NOT proceed until the user explicitly approves.
 
-The ideation must be approved before proceeding. Do NOT proceed until the user sends a message explicitly approving. Wait for explicit approval.
+### Next step
 
-### Proposed next steps
+**STOP** — do NOT propose `/create-prd` or any other pipeline workflow. The only valid next step is:
 
-**Mandatory next step**: Run `/audit-ambiguity ideation` for all inputs, regardless of input type. Even a rich document can have gaps the agent missed. The audit is cheap; the cost of a gap propagating to architecture is high. Do not propose `/create-prd` until `/audit-ambiguity ideation` has run.
+- `/audit-ambiguity ideation` — mandatory coverage verification before `/create-prd` can begin.

package/template/.agent/workflows/ideate.md

@@ -33,22 +33,7 @@ shards: [ideate-extract, ideate-discover, ideate-validate]
 
 After input classification, the user chooses how much involvement they want. All tiers are available for all input types — the pipeline recommends a default but the user picks.
 
-| Tier | Gate behavior | Default for |
-|------|---------------|-------------|
-| 🤖 **Auto** | Pipeline uses Deep Think to self-interview at every gate. User reviews compiled output at the end. | — |
-| 🤝 **Hybrid** | Structural/mechanical gates auto-confirm. Product decisions (personas, MoSCoW, competitive positioning, constraints) pause for user. | Rich doc, thin doc, chat transcript |
-| 💬 **Interactive** | Every gate pauses for explicit user confirmation. | Verbal / one-liner |
-
-**Full matrix — Input × Tier:**
-
-| Input Type | 🤖 Auto | 🤝 Hybrid | 💬 Interactive |
-|---|---|---|---|
-| Rich document | AI extracts + self-interviews gaps | AI extracts, pauses for product calls | AI extracts, pauses at every gate |
-| Thin document | AI expands all domains independently | AI expands, pauses for product calls | AI expands with user at every step |
-| Chat transcript | AI filters noise + self-interviews | AI filters, pauses for product calls | AI filters with user validation |
-| Verbal / one-liner | AI generates vision from scratch via Deep Think | AI generates, pauses for product calls | Full traditional interview |
-
-**Quality guarantee**: All input types and engagement tiers produce the **same output quality** using the same fractal structure. The ideation output from an Auto one-liner is structurally identical to an Interactive rich document. Every node has an index, CX file, and children. Every feature has a Role Lens. Only the amount of human involvement differs.
+Read the engagement tier protocol (`.agent/skills/prd-templates/references/engagement-tier-protocol.md`) — apply the tier behavior for ideation decisions. The user chooses their tier after input classification. All tiers produce the **same output quality** using the same fractal structure — only the amount of human involvement differs.
 
 Transform a raw idea into comprehensive, structured ideation output through exhaustive recursive exploration with the Deep Think protocol.
 
@@ -66,7 +51,32 @@ Transform a raw idea into comprehensive, structured ideation output through exha
 Check whether `docs/plans/ideation/ideation-index.md` already exists.
 
 - **If it does NOT exist** → This is a fresh ideation. Proceed to Shard 1.
-- **If it DOES exist** → **STOP**. Present to the user:
+- **If it exists but is corrupt** (file size < 100 bytes OR missing `## Structural Classification` section) → Treat as fresh. Warn: "Found a corrupt ideation-index.md — treating as a fresh start." Delete the corrupt file and proceed to Shard 1.
+- **If it DOES exist and is valid** → Check for downstream artifacts before offering overwrite:
+
+### 0.1. Downstream cascade check
+
+Scan for downstream pipeline output:
+- `docs/plans/*-architecture-design.md`
+- `docs/plans/ia/` (any `.md` files besides `index.md`)
+- `docs/plans/be/` (any `.md` files besides `index.md`)
+- `docs/plans/fe/` (any `.md` files besides `index.md`)
+
+**If downstream artifacts exist** → **STOP**. Present:
+
+> ⚠️ **Ideation output AND downstream specs exist.**
+>
+> Re-running `/ideate` will invalidate **all** downstream work:
+> - Architecture design document
+> - IA specs (N shards)
+> - BE specs (N specs)
+> - FE specs (N specs)
+>
+> You would need to re-run the entire pipeline from `/create-prd` forward.
+>
+> **Overwrite** (start fresh, accept cascade invalidation) or **Abort**?
+
+**If no downstream artifacts exist** → Present the standard overwrite prompt:
 
 > ⚠️ **Ideation output already exists** at `docs/plans/ideation/ideation-index.md`.
 >
@@ -109,7 +119,10 @@ Explores constraints, success metrics, and competitive positioning. Runs leaf-no
 > The quality self-check and review request are handled by `ideate-validate.md` (Step 12).
 > The parent does not duplicate shard-level quality gates.
 
-### Proposed next steps
+### Next step
+
+**STOP** — do NOT propose `/create-prd` or any other pipeline workflow. The only valid next step is:
 
-**Mandatory next step**: Run `/audit-ambiguity ideation` for all inputs, regardless of input type. Even a rich document can have gaps the agent missed. The audit is cheap; the cost of a gap propagating to architecture is high. Do not propose `/create-prd` until `/audit-ambiguity ideation` has run.
+- `/audit-ambiguity ideation` — mandatory coverage verification before `/create-prd` can begin.
 
+> If the user wants to pause, save progress and note where to resume. When resuming, the next step remains `/audit-ambiguity ideation`.

package/template/.agent/workflows/implement-slice-setup.md

@@ -17,7 +17,7 @@ pipeline:
 
 Check progress state, load skills, read the slice, detect parallel mode, and write the contract.
 
-**Prerequisite**: Phase plan must exist with slice acceptance criteria. If not, tell the user to run `/plan-phase` first.
+**Prerequisite**: Phase plan must exist with slice acceptance criteria. If the phase plan file does not exist → **STOP**: tell the user to run `/plan-phase` first.
 
 ---
 
@@ -64,7 +64,7 @@ Read `.agent/skills/prd-templates/references/skill-loading-protocol.md` and load
 
 Use `find-skills` to discover a test framework skill if needed.
 
-If this slice introduces a new dependency, read `.agent/workflows/bootstrap-agents.md` and execute with the new value.
+If this slice introduces a new dependency, read `.agent/workflows/bootstrap-agents.md` and execute with the new value. **HARD GATE**: Follow the bootstrap verification protocol (`.agent/skills/prd-templates/references/bootstrap-verification-protocol.md`).
 
 ---
 
@@ -72,6 +72,16 @@ If this slice introduces a new dependency, read `.agent/workflows/bootstrap-agen
 
 Read the slice's acceptance criteria from the phase plan.
 
+## 1.25. Load spec context
+
+For each acceptance criterion, trace its spec citation (e.g., `[BE §3.2]`, `[FE §LoginForm]`) back to the source spec:
+
+1. Read the full §section from every cited BE spec — not just the contract shape, but the error handling, edge cases, access control rules, rate limits, and concurrency notes
+2. Read the full §section from every cited FE spec — component props, states, interactions, responsive behavior, accessibility rules
+3. Read any IA shard sections cited by the BE spec's Source Map — especially `## Edge Cases` and `## Access Control`
+
+This context persists throughout the TDD cycle. The acceptance criteria define WHAT to test; the spec context defines HOW DEEP to test it.
+
 ## 1.5. Check for parallel mode
 
 Scan the slice's tasks for surface tags (`BE`, `FE`, `QA`):
@@ -108,8 +118,8 @@ Load the Languages skill(s) from this slice's surface row per the skill loading
 
 Define request/response shapes as {{CONTRACT_LIBRARY}} schemas in the contracts directory (see `.agent/instructions/structure.md`). This is the source of truth.
 
-### Propose next step
+### Next step
 
-Contract written. Next: `/implement-slice-tdd` for the Red→Green→Refactor cycle.
+**STOP** — do NOT proceed to any other workflow. The only valid next step is `/implement-slice-tdd`.
 
-> If invoked standalone, surface via `notify_user`.
+> If invoked standalone, surface via `notify_user` and wait for user confirmation.