qualia-framework 2.1.3 → 2.1.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -25,6 +25,27 @@ Parse JSON for: `researcher_model`, `synthesizer_model`, `roadmapper_model`, `co
25
25
  git init
26
26
  ```
27
27
 
28
+ ## 1.5. Scan Uploaded Files for Tech Stack Conflicts
29
+
30
+ **Before asking any questions**, check if the user placed files in the project directory:
31
+
32
+ ```bash
33
+ ls *.pdf *.docx *.doc *.txt *.md *.png *.jpg *.jpeg 2>/dev/null | head -10
34
+ ```
35
+
36
+ If files are found, read them. If ANY file mentions a tech stack that differs from the Qualia standard (Next.js + React + TypeScript + Supabase + Vercel), you MUST ask the user:
37
+
38
+ Use AskUserQuestion:
39
+ - header: "Tech Stack"
40
+ - question: "The uploaded document mentions {detected stack}. Which stack should we use?"
41
+ - options:
42
+ - "Qualia stack (Recommended)" — Next.js + React + TypeScript + Supabase + Vercel. This is what we build with.
43
+ - "Document stack" — Use {detected stack} as specified in the document. The workflow adapts.
44
+
45
+ **Never silently adopt a non-Qualia stack from uploaded documents.** Always confirm first.
46
+
47
+ If no files found or no stack conflict: skip silently.
48
+
28
49
  ## 2. Brownfield Offer
29
50
 
30
51
  **If `needs_codebase_map` is true** (from init — existing code detected but no codebase map):
@@ -14,6 +14,22 @@ Initialize a new project with the full Qualia qualia-engine workflow: questionin
14
14
 
15
15
  ## Process
16
16
 
17
+ ### 0. Scan for Uploaded Files
18
+
19
+ Before asking any questions, check if the user uploaded or placed files in the project directory:
20
+
21
+ ```bash
22
+ ls *.pdf *.docx *.doc *.txt *.md *.png *.jpg *.jpeg 2>/dev/null | head -10
23
+ ```
24
+
25
+ If files are found, read them. If ANY file mentions a tech stack (WordPress, Firebase, PHP, Laravel, React Native, Flutter, etc.) that differs from the Qualia standard stack (Next.js + Supabase + Vercel), you MUST ask:
26
+
27
+ > "The document mentions {tech stack from doc}. Do you want to:
28
+ > 1. **Use the Qualia stack** (Next.js + React + TypeScript + Supabase + Vercel) — recommended, this is what we build with
29
+ > 2. **Use the stack from the document** ({tech stack}) — I'll adapt the workflow accordingly"
30
+
31
+ Never silently adopt a non-Qualia stack from uploaded documents. Always confirm.
32
+
17
33
  ### 1. Questioning Phase
18
34
 
19
35
  Run the questioning workflow to understand the project deeply before any planning.
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "qualia-framework",
3
- "version": "2.1.3",
3
+ "version": "2.1.4",
4
4
  "description": "Qualia Solutions — Claude Code Framework",
5
5
  "bin": {
6
6
  "qualia-framework": "./bin/cli.js"
@@ -39,12 +39,12 @@ Collaborator, not executor. Speak up about bugs, simpler approaches, bad archite
39
39
  Be honest. Default to action. Never speculate on unread code. Say when blocked.
40
40
 
41
41
  ## Workflow
42
- - Session start: `pwd` > identify project > read local CLAUDE.md
42
+ - **MANDATORY FIRST ACTION**: On every session start, invoke the `qualia-start` skill before doing anything else. This is non-negotiable — do not wait for user input, do not skip it, do not just acknowledge the hook message. Actually invoke the skill using the Skill tool.
43
43
  - Subagents (Sonnet) for research. Opus for complex reasoning.
44
44
  - `/compact` at 60%. `/clear` between tasks. `/learn` after mistakes.
45
45
 
46
- ## Qualia Mode (activate with /qualia-start)
47
- When Qualia mode is active, these behaviors apply to ALL interactions:
46
+ ## Qualia Mode (always active)
47
+ These behaviors apply to ALL interactions:
48
48
  - **Frontend guard:** Read .planning/DESIGN.md before any frontend file changes
49
49
  - **Deploy guard:** Check .planning/REVIEW.md freshness before any deploy command (run /qualia-review to generate)
50
50
  - **Intent verification:** Confirm before modifying 3+ files in one response
@@ -21,6 +21,15 @@
21
21
  },
22
22
  "hooks": {
23
23
  "PreToolUse": [
24
+ {
25
+ "matcher": "Skill",
26
+ "hooks": [
27
+ {
28
+ "type": "command",
29
+ "command": "~/.claude/hooks/skill-announce.sh"
30
+ }
31
+ ]
32
+ },
24
33
  {
25
34
  "matcher": "Edit",
26
35
  "hooks": [
@@ -80,6 +89,17 @@
80
89
  ]
81
90
  }
82
91
  ],
92
+ "PostToolUseFailure": [
93
+ {
94
+ "matcher": "",
95
+ "hooks": [
96
+ {
97
+ "type": "command",
98
+ "command": "~/.claude/hooks/tool-error-announce.sh"
99
+ }
100
+ ]
101
+ }
102
+ ],
83
103
  "PreCompact": [
84
104
  {
85
105
  "matcher": "",
@@ -91,6 +111,17 @@
91
111
  ]
92
112
  }
93
113
  ],
114
+ "PostCompact": [
115
+ {
116
+ "matcher": "",
117
+ "hooks": [
118
+ {
119
+ "type": "command",
120
+ "command": "~/.claude/hooks/session-context-loader.sh"
121
+ }
122
+ ]
123
+ }
124
+ ],
94
125
  "SessionStart": [
95
126
  {
96
127
  "matcher": "",