Intent Integrity Kit Plan

Generate design artifacts from the feature specification using the plan template.

User Input

$ARGUMENTS

You MUST consider the user input before proceeding (if not empty).

Constitution Loading

Load constitution per constitution-loading.md (enforcement mode — extract rules, declare hard gate, halt on violations).

Prerequisites Check

1. Run prerequisites check:

   bash .tessl/tiles/tessl-labs/intent-integrity-kit/skills/iikit-03-plan/scripts/bash/check-prerequisites.sh --phase 03 --json

   Windows: pwsh .tessl/tiles/tessl-labs/intent-integrity-kit/skills/iikit-03-plan/scripts/powershell/check-prerequisites.ps1 -Phase 03 -Json

2. Parse JSON for FEATURE_SPEC, IMPL_PLAN, FEATURE_DIR, BRANCH. If missing spec.md: ERROR.

3. If JSON contains needs_selection: true: present the features array as a numbered table (name and stage columns). Follow the options presentation pattern in conversation-guide.md. After user selects, run:

   bash .tessl/tiles/tessl-labs/intent-integrity-kit/skills/iikit-03-plan/scripts/bash/set-active-feature.sh --json <selection>

   Windows: pwsh .tessl/tiles/tessl-labs/intent-integrity-kit/skills/iikit-03-plan/scripts/powershell/set-active-feature.ps1 -Json <selection>

   Then re-run the prerequisites check from step 1.

Spec Quality Gate

Before planning, validate spec.md:

1. Requirements: count FR-XXX patterns (ERROR if 0, WARNING if <3)
2. Measurable criteria: scan for numeric values, percentages, time measurements (WARNING if none)
3. Unresolved clarifications: search for [NEEDS CLARIFICATION] — ask whether to proceed with assumptions
4. User story coverage: verify each story has acceptance scenarios
5. Cross-references: check for orphan requirements not linked to stories

Report quality score per formatting-guide.md (Spec Quality section). If score < 6: recommend /iikit-02-clarify first.

Execution Flow

1. Fill Technical Context

Using the plan template, define: Language/Version, Primary Dependencies, Storage, Testing, Target Platform, Project Type, Performance Goals, Constraints, Scale/Scope. Mark unknowns as "NEEDS CLARIFICATION".

When Tessl eval results are available for candidate technologies, include eval scores in the decision rationale in research.md. Higher eval scores indicate better-validated tiles and should factor into technology selection when choosing between alternatives.

2. Tessl Tile Discovery

If Tessl is installed, discover and install tiles for all technologies. See tessl-tile-discovery.md for the full procedure.

3. Research & Resolve Unknowns

For each NEEDS CLARIFICATION item and dependency: research, document findings in research.md with decision, rationale, and alternatives considered. Include Tessl Tiles section if applicable.

4. Design & Contracts

Prerequisites: research.md complete

1. Extract entities from spec -> data-model.md (fields, relationships, validation, state transitions)
2. Generate API contracts from functional requirements -> contracts/
3. Create quickstart.md with test scenarios
4. Update agent context:

   bash .tessl/tiles/tessl-labs/intent-integrity-kit/skills/iikit-03-plan/scripts/bash/update-agent-context.sh claude

   Windows: pwsh .tessl/tiles/tessl-labs/intent-integrity-kit/skills/iikit-03-plan/scripts/powershell/update-agent-context.ps1 -AgentType claude

5. Pre-compute Dashboard Data

After the plan is complete, write pre-computed data to .specify/context.json for static dashboard generation. Use jq to merge into the existing file (create if missing).

5a. Architecture Node Classifications

If plan.md contains an architecture diagram (ASCII box-drawing), classify each named component as one of: client, server, storage, external.

Write to .specify/context.json under planview.nodeClassifications:

# Read existing or start fresh
CONTEXT_FILE=".specify/context.json"
[[ -f "$CONTEXT_FILE" ]] || echo '{}' > "$CONTEXT_FILE"

# Merge node classifications (replace example with actual nodes from the plan diagram)
jq --argjson nodes '{
  "Browser SPA": "client",
  "API Gateway": "server",
  "PostgreSQL": "storage",
  "Stripe API": "external"
}' '.planview.nodeClassifications = $nodes' "$CONTEXT_FILE" > "$CONTEXT_FILE.tmp" && mv "$CONTEXT_FILE.tmp" "$CONTEXT_FILE"

Classification rules:

• client: browsers, CLIs, mobile apps, desktop apps — anything that initiates requests
• server: APIs, gateways, workers, middleware, backend services — anything that processes requests
• storage: databases, caches, queues, file stores, object storage — anything that persists data
• external: third-party APIs, SaaS services, payment providers — anything outside the project boundary

If no architecture diagram exists in the plan, skip this step.

5b. Tessl Eval Scores

If Tessl tiles were installed in step 2, collect eval scores from the fetch-tile-evals.sh outputs and write a summary to context.json:

# Merge eval scores (replace example with actual tile names and scores from step 2)
jq --argjson evals '{
  "workspace/tile-name": {"score": 85, "pct": 85, "scenarios": 3, "scored_at": "2026-01-15T10:00:00Z"}
}' '.planview.evalScores = $evals' "$CONTEXT_FILE" > "$CONTEXT_FILE.tmp" && mv "$CONTEXT_FILE.tmp" "$CONTEXT_FILE"

Use the JSON output from each fetch-tile-evals.sh --json call (already run in step 2 via tessl-tile-discovery.md). Extract score, pct, scenarios, and scored_at fields for each tile.

If no Tessl tiles were installed, skip this step.

6. Constitution Check (Post-Design)

Re-validate all technical decisions against constitutional principles. On violation: STOP, state violation, suggest compliant alternative.

7. Phase Separation Validation

Scan plan for governance content per phase-separation-rules.md (Plan section). Auto-fix by replacing with constitution references, re-validate.

Output Validation

Before writing any artifact: review against each constitutional principle. On violation: STOP with explanation and alternative.

Report

Output: branch name, plan path, generated artifacts (research.md, data-model.md, contracts/*, quickstart.md), agent file update status, Tessl integration status (tiles installed, skills available, technologies without tiles, eval results saved), dashboard pre-computed data status (node classifications written, eval scores written).

Semantic Diff on Re-run

If plan.md exists: compare tech stack, architecture, dependencies. Show diff per formatting-guide.md (Semantic Diff section) with downstream impact. Flag breaking changes.

Next Steps

You MUST read model-recommendations.md, check the expiration date (refresh via web search if expired), detect the agent via env vars, and include a model switch tip in the output below if the next phase needs a different model tier.

Plan complete! Next steps:
- /iikit-04-checklist - (Recommended) Generate quality checklists
- /iikit-05-testify - (REQUIRED by constitution) Generate test specifications [if TDD mandatory]
- /iikit-05-testify - (Optional) Generate test specifications for TDD [if TDD not mandatory]
- /iikit-06-tasks - Generate task breakdown from plan
Tip: <model switch suggestion if tier mismatch, omit if already on the right model>
- Dashboard: file://$(pwd)/.specify/dashboard.html (resolve the path)