chati-dev 4.0.3 → 4.0.4

package/framework/agents/build/dev.md

@@ -54,7 +54,32 @@ Implement each task from the approved task breakdown with high quality, followin
 ```
 For each task:
   1. Announce: "Starting T{X}: {title} — implementing now..."
-  2. Read task details and acceptance criteria
+  1.5. Pre-Flight Spec Check (BEFORE any code):
+     a. Validate Given-When-Then criteria:
+        → Are all criteria specific and measurable? (not "works correctly", "looks good")
+        → If ANY criterion is vague/untestable → STOP. Escalate G01: "{criterion} is ambiguous."
+        → If task has NO criteria → STOP. Escalate G01: "T{X} has no testable acceptance criteria."
+     b. Check dependencies:
+        → For each T{x}.{y} in Dependencies: is it marked complete in tasks.md?
+        → If dependency incomplete → STOP. Escalate: "Cannot start T{X}: T{dep} not yet complete."
+     c. Set implementation strategy from task size:
+        → XS/S: standard flow — proceed to Step 2
+        → M (2-4h): output brief implementation outline (files + approach), then proceed
+        → L (4-8h): output full implementation plan, ask "[ready/clarify/skip]", wait for user
+     RULE: NEVER write code before Step 1.5 passes. If spec is unclear, fix the spec first.
+     OUTPUT: "Spec check passed. T{X} is {size} — {strategy}."
+  2. Read task details, acceptance criteria, and verify architectural alignment:
+     Before coding, cross-check against chati.dev/artifacts/3-Architecture/architecture.md:
+     - API tasks → endpoint pattern, response format, error handling contract match Section 4?
+     - Database tasks → table/column names, relationships match Section 5 (Data Model)?
+     - Auth tasks → auth approach matches Section 6 (Authentication)?
+     - New modules/components → file structure matches Section 3 (System Components)?
+     If conflict detected → STOP. Escalate G05:
+     "T{X} conflicts with architecture decision: {specific conflict}.
+      Implement per architecture.md or per task spec?"
+     RULE: architecture.md is the source of truth. NEVER implement against architectural decisions.
+     RULE: Do NOT silently reconcile conflicts — always surface them.
+     If no architecture.md present → proceed with best practices, note in handoff.
   3. Implement code
      -> Output: "Implementation done. Running self-critique (5.5)..."
   4. Run self-critique (Step 5.5) — 1 fix pass, then proceed
@@ -89,7 +114,17 @@ WHILE tasks_pending:
 
   FOR attempt IN 1..3:
     1. Read task details and acceptance criteria
-    2. Implement code
+    1.5. Pre-Flight Spec Check:
+         → If ANY criterion is vague/untestable → mark task blocked (G01), skip to next task
+         → If task has no Given-When-Then criteria → mark task blocked (G01), skip to next
+         → If dependency not complete → mark task blocked, skip to next independent task
+         → M-size tasks: output implementation outline before coding
+         → L-size tasks: output full implementation plan before coding (no user confirmation)
+    2. Verify architectural alignment (cross-check architecture.md before coding):
+         → API/DB/Auth/module tasks: verify patterns match architecture.md
+         → If conflict → mark task blocked (G05), skip to next task
+         → If no architecture.md → proceed with best practices
+    3. Implement code
        -> Output: "T{X} implementation done. Self-critique (5.5)..."
     3. Run self-critique (Step 5.5) — 1 fix pass, then proceed
        -> Output: "T{X} critique done. Running tests..."
@@ -282,9 +317,11 @@ Criteria:
 6. No lint errors
 7. Self-critique (5.5 + 6.5) completed
 8. No blockers remaining
+9. Pre-flight spec check passed: criteria were specific and testable before coding started
+10. Architectural alignment verified: no G05 conflicts, or G05 explicitly resolved before coding
 
 Score = criteria met / total criteria
-Threshold: >= 95% per task
+Threshold: >= 95% per task (minimum 9/10)
 ```
 
 ---
@@ -325,7 +362,7 @@ agents:
   dev:
     status: in_progress | completed
     score: {average across all tasks}
-    criteria_count: {total criteria across all tasks}
+    criteria_count: 10
     completed_at: "{timestamp when all tasks done}"
 ```
 

package/framework/config.yaml

@@ -1,8 +1,8 @@
 # chati.dev Configuration
-version: "4.0.3"
+version: "4.0.4"
 installed_at: "2026-02-07T10:00:00Z"
-updated_at: "2026-03-21T00:00:00Z"
-installer_version: "4.0.3"
+updated_at: "2026-03-22T00:00:00Z"
+installer_version: "4.0.4"
 project_type: greenfield
 language: en
 ides: [claude-code]

package/framework/domains/agents/dev.yaml

@@ -48,6 +48,19 @@ rules:
     text: "Iterate on implementation until acceptance criteria are met."
     priority: normal
 
+  - id: dev-spec-intake
+    text: "MUST validate task spec BEFORE writing code: verify all Given-When-Then criteria are specific and testable, check all task dependencies are complete. Escalate G01 immediately for ambiguous criteria, G02 for conflicting criteria — NEVER start coding with unclear spec."
+    priority: critical
+
+  - id: dev-size-strategy
+    text: "MUST adapt implementation approach by task size: XS/S = standard flow; M (2-4h) = output implementation outline before coding; L (4-8h) = output implementation plan and ask user confirmation before starting (autonomous mode: output plan and proceed without confirmation)."
+    priority: high
+
+  - id: dev-arch-coherence
+    text: "MUST verify implementation aligns with architecture.md before coding any API endpoints, database schemas, auth flows, or new modules. Escalate G05 immediately if conflict detected. NEVER implement against architectural decisions — always surface conflicts to user."
+    priority: high
+
   # Provider Preference (v3.0.0)
-  - dev-provider: "Default provider: claude. Can run on gemini for large codebase tasks (1M context window)."
+  - id: dev-provider
+    text: "Default provider: claude. Can run on gemini for large codebase tasks (1M context window)."
     priority: normal

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "chati-dev",
-  "version": "4.0.3",
+  "version": "4.0.4",
   "description": "AI-Powered Multi-Agent Orchestration System — Structured vibe coding for Full Stack Development",
   "type": "module",
   "bin": {