ai-consultants 3.5.0 → 4.0.1

package/.claude/commands/ai-consultants:consult.md

@@ -1,5 +1,5 @@
 ---
-description: Consult AI experts (Gemini, Codex, Mistral, Cursor, Kimi, Qwen, GLM, Grok, DeepSeek, MiniMax) for coding questions. Use when weighing trade-offs, comparing approaches, or wanting multiple expert perspectives on non-trivial decisions.
+description: Consult AI experts (Gemini, Codex, Mistral, Kimi, Qwen, GLM, Grok, DeepSeek, MiniMax) for coding questions. Use when weighing trade-offs, comparing approaches, or wanting multiple expert perspectives on non-trivial decisions.
 argument-hint: <question> [file1] [file2] ...
 allowed-tools: Bash Read Glob Grep
 ---
@@ -14,6 +14,11 @@ Query multiple AI models for expert opinions on a coding question.
 
 1. If no question provided in $ARGUMENTS, ask the user what they want to consult about.
 
+   Separate supported consultation options from the question before running the
+   panel. In particular, preserve `--preset <name>` and `--strategy <name>` as
+   CLI arguments placed before the question; they are not part of the question
+   text. Reject an incomplete option instead of guessing its value.
+
 2. **File context handling**: Identify which strings in $ARGUMENTS refer to files the user wants the consultants to look at. Use your judgment — verify existence via `Glob` or `Bash ls` when uncertain. This replaces fragile regex detection: a `Makefile`, `Dockerfile`, dotfile, or path without a common extension is still a file; an URL or regex pattern in the question text is not.
 
    For each file you identify:
@@ -25,22 +30,35 @@ Query multiple AI models for expert opinions on a coding question.
 
    **Note (Claude only):** Do not use your `Read` tool to dump file contents into the query. `build_context.sh` handles file reading itself, without the `N\t` line-number prefix that `Read` adds.
 
-3. **Run the consultation**:
+3. **Write the question safely and run the consultation**:
+   - With Bash, create a query file using `mktemp`, set mode 600, and write the
+     exact question text (with file paths and consultation options removed)
+     through a **single-quoted heredoc delimiter that does not occur as its own
+     line in the question**. Never pass raw user text as a shell argument or
+     through an unquoted heredoc.
+   - Shell-quote every context-file path as one argument.
+   - Install a shell trap that removes only the exact query file when the Bash
+     invocation exits; question text must not persist in the temp directory.
+   - Capture `caller_root="$PWD"` before changing directories. Pass it through
+     `--context-root` so regular files from the user's project are staged under
+     a private temp directory without widening the runtime filesystem allowlist.
+
    ```bash
-   cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && INVOKING_AGENT=claude ./scripts/consult_all.sh '<question>' <file1[@TAG]> <file2[@TAG]> ...
+   caller_root="$PWD"; cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && INVOKING_AGENT=claude ./scripts/consult_all.sh [--preset <name>] [--strategy <name>] --context-root "$caller_root" --query-file "<private-question-file>" "<file1[@TAG]>" "<file2[@TAG]>" ...
    ```
-   - `<question>` is the question text with file paths stripped. Use single quotes to prevent shell expansion.
    - Each `<fileN>` is a file path, optionally suffixed with `@PRIMARY` or `@CONTEXT`.
-   - **If the question text exceeds ~8KB or contains awkward quoting**, write it to a tmpfile and use `--query-file <path>` instead:
-     ```bash
-     cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && INVOKING_AGENT=claude ./scripts/consult_all.sh --query-file /tmp/question.txt <file1[@TAG]> ...
-     ```
    - **Capture the last line of stdout** — it contains the output directory path (e.g., `/tmp/ai_consultations/20260315_143022_12345/`).
+   - Keep `INVOKING_AGENT=claude` exactly as shown. Never invoke
+     `query_claude.sh` directly and never remove self-exclusion to increase the
+     panel count.
 
 4. **Read the results** from the output directory captured in step 3:
    - Read `<output_dir>/report.md` using the **Read** tool
    - Read `<output_dir>/synthesis.json` using the **Read** tool (may not exist if synthesis was skipped or failed)
    - If neither exists, use the **Glob** tool with pattern `<output_dir>/*.json` to discover individual consultant response files, then read each one
+   - `claude.json` must not exist and `.synthesis_provider` must not be
+     `claude`. If either appears, report a self-exclusion failure instead of
+     presenting the run as valid.
 
 5. **Present the results** using this template. The panel's value is *coverage* — the
    union of distinct points across diverse models, including what only one raised — not a
@@ -87,5 +105,7 @@ Query multiple AI models for expert opinions on a coding question.
 
 For follow-up questions on the same consultation:
 ```bash
-cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && INVOKING_AGENT=claude ./scripts/followup.sh '<follow-up question>'
+cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && INVOKING_AGENT=claude ./scripts/followup.sh --query-file "<private-follow-up-file>"
 ```
+Create and clean the follow-up file with the same private heredoc/trap contract.
+Never target Claude from a Claude-hosted follow-up.

package/.claude/commands/ai-consultants:help.md

@@ -19,9 +19,13 @@ All skill details (consultants, personas, presets, strategies, features, configu
 ```
 /ai-consultants:consult "How to optimize this SQL query?"
 /ai-consultants:consult "Review this code" src/utils.ts
-/ai-consultants:debate "Microservices or monolith?"
+/ai-consultants:consult --preset max_quality "Microservices or monolith?"
 ```
 
+When Claude invokes the panel, the command sets `INVOKING_AGENT=claude` so
+Claude is excluded from both consultation and synthesis while Codex and the
+other enabled providers remain available.
+
 Configuration (presets, strategies, features, personas, API keys) can be managed via natural language — just ask.
 
 Troubleshooting: run `./scripts/doctor.sh` or `./scripts/doctor.sh --fix`.

package/.env.example

@@ -65,13 +65,12 @@
 # ENABLED CONSULTANTS
 # =============================================================================
 # Set to "false" to disable a consultant you haven't configured.
-# MINIMUM 2 consultants must be enabled for comparison/voting to work.
+# MINIMUM 2 consultants must be enabled for a useful coverage panel.
 
 # CLI-based consultants (default: enabled)
 ENABLE_GEMINI=true
 ENABLE_CODEX=true
 ENABLE_MISTRAL=true
-ENABLE_CURSOR=true
 ENABLE_KIMI=true
 ENABLE_QWEN3=true
 ENABLE_GROK=true
@@ -113,17 +112,18 @@ ENABLE_DEEPSEEK=false
 #   https://token-plan.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1/chat/completions
 
 # =============================================================================
-# MODEL CONFIGURATION (June 2026 premium defaults)
+# MODEL CONFIGURATION (August 2026 premium defaults)
 # =============================================================================
 
 # Gemini - The Architect (CLI mode via Antigravity CLI: agy)
-GEMINI_MODEL="Gemini 3.1 Pro (High)"
+GEMINI_MODEL="Gemini 3.7 Flash (High)"
 GEMINI_TIMEOUT=180
 GEMINI_CMD=agy
 # API-mode model ID (only used when GEMINI_USE_API=true)
 GEMINI_API_MODEL=gemini-3.1-pro-preview
-# Authenticated opt-ins: use Gemini 3.7 Flash (High/Low) for GEMINI_MODEL or
-# gemini-3.7-flash for GEMINI_API_MODEL.
+# Gemini 3.1 Pro and 3.6 Flash remain valid CLI pins. API mode stays on the
+# separately verified 3.1 Pro ID; gemini-3.7-flash is an API-only opt-in until
+# that transport completes its own authenticated smoke.
 # Reasoning effort for the agy CLI. Passed as --effort <value>. Accepts only
 # low|medium|high; anything else is rejected by the provider and surfaces as a
 # diagnosed error (provider owns the enum — same stance as QWEN3_REASONING_EFFORT).
@@ -131,6 +131,7 @@ GEMINI_API_MODEL=gemini-3.1-pro-preview
 
 # Codex - The Pragmatist
 CODEX_MODEL=gpt-5.6-sol
+CODEX_API_MAX_TOKENS=4096
 CODEX_TIMEOUT=180
 CODEX_CMD=codex
 # Reasoning effort for the codex CLI (and API mode). Passed as
@@ -143,15 +144,12 @@ CODEX_CMD=codex
 # Large 3, and Small 4 API IDs are documented opt-ins until an API smoke passes.
 MISTRAL_MODEL=mistral-large-3
 MISTRAL_CLI_MODEL=mistral-medium-3.5
+MISTRAL_API_MAX_TOKENS=4096
 # Official API opt-ins: use mistral-medium-3-5, mistral-large-2512, or
 # mistral-small-2603 for MISTRAL_MODEL after an authenticated API smoke.
 MISTRAL_TIMEOUT=180
 MISTRAL_CMD=vibe
-
-# Cursor - The Integrator
-CURSOR_MODEL=composer-2.5
-CURSOR_TIMEOUT=180
-CURSOR_CMD=cursor-agent
+MISTRAL_MAX_TURNS=4
 
 # Kimi - The Eastern Sage (K3 via Kimi Code CLI)
 KIMI_MODEL=kimi-code/k3
@@ -174,6 +172,8 @@ CLAUDE_API_MAX_TOKENS=16384
 # Qwen3 - The Analyst (CLI/API switchable v2.7)
 QWEN3_MODEL=qwen3.7-max
 QWEN3_TIMEOUT=180
+QWEN3_MAX_QUALITY_TIMEOUT=600
+QWEN3_API_MAX_TOKENS=16384
 QWEN3_CMD=qwen
 # Reasoning effort. API mode ONLY - the qwen CLI has no effort flag, so in CLI
 # mode this is ignored (with a warning); set model.reasoningEffort in
@@ -196,24 +196,36 @@ QWEN3_CMD=qwen
 # GLM - The Methodologist
 GLM_MODEL=glm-5.3
 GLM_TIMEOUT=180
+GLM_API_MAX_TOKENS=16384
 GLM_API_URL=https://api.z.ai/api/coding/paas/v4/chat/completions
+# GLM_REASONING_EFFORT=  # max_quality sets max
 
 # Grok Build CLI + xAI API fallback - The Provocateur
 GROK_CMD=grok
 GROK_MODEL=grok-4.6
 GROK_TIMEOUT=180
+GROK_MAX_TURNS=4
+GROK_API_MAX_TOKENS=4096
 GROK_API_URL=https://api.x.ai/v1/chat/completions
+# GROK_REASONING_EFFORT= # max_quality sets xhigh (Grok Build maximum)
 
 # DeepSeek - The Code Specialist
 DEEPSEEK_MODEL=deepseek-v4-pro
 DEEPSEEK_TIMEOUT=180
+DEEPSEEK_MAX_QUALITY_TIMEOUT=600
+DEEPSEEK_API_MAX_TOKENS=16384
 DEEPSEEK_API_URL=https://api.deepseek.com/v1/chat/completions
+# DEEPSEEK_REASONING_EFFORT= # max_quality sets max
 
 # MiniMax - The Pragmatic Optimizer (CLI via mmx by default, v2.21; API opt-in)
 MINIMAX_CMD=mmx
 MINIMAX_MODEL=MiniMax-M2.7
 # max_quality selects the smoke-tested MiniMax-M3 target.
 MINIMAX_TIMEOUT=180
+MINIMAX_API_MAX_TOKENS=4096
+MINIMAX_MAX_TOKENS=4096
+MINIMAX_MAX_QUALITY_TOKENS=16384
+MAX_QUALITY_API_MAX_TOKENS=16384
 MINIMAX_API_URL=https://api.minimax.io/v1/chat/completions
 
 # Wire format for a consultant's API mode. Honored since v2.25 (previously
@@ -249,6 +261,9 @@ SYNTHESIS_CMD=claude
 SYNTHESIS_MODEL=
 SYNTHESIS_EXTRACT_FIELDS=true
 SYNTH_MAX=10
+SYNTH_DETAIL_MAX_CHARS=4000
+SYNTHESIS_TIMEOUT=240
+SYNTHESIS_TOTAL_TIMEOUT=480
 
 # Health gate and quorum: validate the live panel before spending on a full run
 ENABLE_HEALTH_GATE=false       # Adds one small ping per selected consultant
@@ -388,7 +403,6 @@ REPORT_MAX_JSON_LINES=50
 # GEMINI_PERSONA_ID=1
 # CODEX_PERSONA_ID=2
 # MISTRAL_PERSONA_ID=3
-# CURSOR_PERSONA_ID=5
 # KIMI_PERSONA_ID=20
 # CLAUDE_PERSONA_ID=18
 # QWEN3_PERSONA_ID=6
@@ -402,8 +416,6 @@ REPORT_MAX_JSON_LINES=50
 # CODEX_PERSONA_NAME=Custom persona name
 # MISTRAL_PERSONA=Custom persona instructions
 # MISTRAL_PERSONA_NAME=Custom persona name
-# CURSOR_PERSONA=Custom persona instructions
-# CURSOR_PERSONA_NAME=Custom persona name
 # KIMI_PERSONA=Custom persona instructions
 # KIMI_PERSONA_NAME=Custom persona name
 # CLAUDE_PERSONA=Custom persona instructions

package/CLAUDE.md

@@ -2,11 +2,11 @@
 
 ## Project Overview
 
-AI Consultants is a multi-model coverage system that queries up to 11 AI consultants (Gemini, Codex, Mistral, Cursor, Kimi, Claude, Qwen3, GLM, Grok, DeepSeek, MiniMax) to obtain diverse perspectives on coding problems.
+AI Consultants is a multi-model coverage system that queries up to 10 AI consultants (Gemini, Codex, Mistral, Kimi, Claude, Qwen3, GLM, Grok, DeepSeek, MiniMax) to obtain diverse perspectives on coding problems.
 
-**Self-Exclusion**: The invoking agent is automatically excluded from the panel to prevent self-consultation. Claude Code won't query Claude, Codex CLI won't query Codex, etc.
+**Self-Exclusion**: The invoking agent is automatically excluded from both the panel and synthesis. Claude Code won't query or synthesize with Claude, Codex CLI won't query or synthesize with Codex, etc.
 
-**Version**: 3.5.0
+**Version**: 4.0.1
 
 ## Distribution
 
@@ -176,25 +176,33 @@ Functions in `lib/common.sh`:
 - `get_self_consultant_name()` - Maps invoking agent to consultant name
 - `should_skip_consultant()` - Returns true if consultant should be excluded
 - `log_self_exclusion_status()` - Debug logging for exclusion
+- `resolve_synthesis_cli()` - Refuses the invoking agent as synthesis provider
+
+Agent hosts must set `INVOKING_AGENT` explicitly and run `consult_all.sh`
+instead of individual adapters. A Claude-hosted run must produce no
+`claude.json`, and its `synthesis_provider` must not be `claude`; the equivalent
+invariant applies to Codex and Gemini. See `SKILL.md` for the full execution
+contract.
 
 ### Configuration Presets
 ```bash
 # Quality Tier Presets (v2.5)
-./scripts/consult_all.sh --preset max_quality "question"  # 8 of 11 consultants + maximum tier + peer review
-./scripts/consult_all.sh --preset medium "question"       # 4 consultants + standard models
+./scripts/consult_all.sh --preset max_quality "question"  # all 10 consultants + maximum tier/max effort
+./scripts/consult_all.sh --preset medium "question"       # 3 consultants + standard models
 ./scripts/consult_all.sh --preset fast "question"         # 2 consultants + economy models
 
 # Use Case Presets
 ./scripts/consult_all.sh --preset minimal "question"    # Gemini + Codex
-./scripts/consult_all.sh --preset balanced "question"   # + Mistral + Cursor
-./scripts/consult_all.sh --preset high-stakes "question" # All + debate
+./scripts/consult_all.sh --preset balanced "question"   # + Mistral
+./scripts/consult_all.sh --preset high-stakes "question" # Broad premium panel
 ```
 
 Presets are defined in `config.sh` via `apply_preset()` function.
 
 ### Synthesis Strategies
 ```bash
-./scripts/consult_all.sh --strategy majority "question"      # Default
+./scripts/consult_all.sh --strategy coverage "question"      # Default: union of distinct points
+./scripts/consult_all.sh --strategy majority "question"      # Blended recommendation
 ./scripts/consult_all.sh --strategy risk_averse "question"   # Conservative
 ./scripts/consult_all.sh --strategy security_first "question" # Security focus
 ./scripts/consult_all.sh --strategy compare_only "question"  # No recommendation
@@ -210,34 +218,6 @@ Strategies are implemented in `synthesize.sh` via `get_strategy_instructions()`.
 ./scripts/doctor.sh --verbose    # Detailed output
 ```
 
-### Panic Button Mode
-Automatically adds rigor when uncertainty detected:
-- Average confidence below `PANIC_CONFIDENCE_THRESHOLD` (default: 5)
-- Uncertainty keywords detected in responses
-
-Configuration in `config.sh`:
-- `ENABLE_PANIC_MODE` - "auto", "always", or "never"
-- `PANIC_CONFIDENCE_THRESHOLD` - Trigger threshold
-- `PANIC_EXTRA_DEBATE_ROUNDS` - Additional debate rounds
-
-### Confidence Intervals
-Functions in `lib/voting.sh`:
-- `calculate_confidence_stddev()` - Returns stddev scaled by 10
-- `calculate_confidence_interval()` - Returns JSON with mean, stddev, interval
-- `has_high_confidence_variance()` - Returns 0 if variance > 2.0
-- `format_confidence_range()` - Returns "7 ± 1.5" format
-
-### Anonymous Peer Review
-```bash
-./scripts/peer_review.sh <responses_dir> <output_dir>
-```
-
-Process:
-1. Anonymize all responses (remove consultant names)
-2. Each consultant ranks and critiques responses
-3. Aggregate peer scores to identify strongest arguments
-4. De-anonymize in final report
-
 ## v2.9 Features
 
 ### Kimi CLI Support
@@ -288,7 +268,7 @@ npm install -g @qwen-code/qwen-code@latest
 ## v2.6 Features
 
 ### CLI/API Mode Switching
-Six consultants can now switch between CLI and API mode: **Gemini, Codex, Claude, Mistral, Qwen3, MiniMax**.
+Seven consultants can switch between CLI and API mode: **Gemini, Codex, Claude, Mistral, Qwen3, Grok, MiniMax**.
 
 When API mode is enabled for an agent, CLI mode is disabled (mutual exclusivity).
 
@@ -325,6 +305,7 @@ New environment variables in `config.sh`:
 | `CLAUDE_USE_API` | false | Use Anthropic API instead of claude CLI |
 | `MISTRAL_USE_API` | false | Use Mistral API instead of vibe CLI |
 | `QWEN3_USE_API` | false | Use DashScope API instead of qwen CLI (v2.7) |
+| `GROK_USE_API` | auto | Use xAI API when Grok Build is unavailable |
 | `MINIMAX_USE_API` | false | Use MiniMax API instead of the mmx CLI (v2.21) |
 | `GEMINI_API_URL` | https://generativelanguage.googleapis.com/v1beta/models | Google AI endpoint |
 | `CODEX_API_URL` | https://api.openai.com/v1/chat/completions | OpenAI endpoint |
@@ -368,10 +349,10 @@ Three normal tiers plus a `maximum` tier are configurable via `apply_model_tier(
 
 | Tier | Description | Example Models |
 |------|-------------|----------------|
-| **premium** | Latest flagship models | claude-opus-5, Gemini 3.1 Pro (High), gpt-5.6-sol |
-| **maximum** | Smoke-tested separate-plan targets used by max_quality | K3-256k, Qwen3.8-Max, MiniMax M3 |
-| **standard** | Good quality at reasonable cost | claude-sonnet-5, Gemini 3.6 Flash (High), gpt-5.6-terra |
-| **economy** | Optimized for speed and low cost | claude-haiku-4-5, Gemini 3.6 Flash (Low), gpt-5.6-luna |
+| **premium** | Latest flagship models | claude-opus-5, Gemini 3.7 Flash (High), gpt-5.6-sol |
+| **maximum** | All 10 consultants; maximum targets and highest provider effort | K3-256k, Qwen3.8-Max, MiniMax M3; Grok xhigh, GLM/DeepSeek max |
+| **standard** | Good quality at reasonable cost | claude-sonnet-5, Gemini 3.7 Flash (High), gpt-5.6-terra |
+| **economy** | Optimized for speed and low cost | claude-haiku-4-5, Gemini 3.7 Flash (Low), gpt-5.6-luna |
 
 **Default models are now premium tier** for maximum quality.
 
@@ -384,7 +365,7 @@ apply_model_tier "standard"  # Set all consultants to standard models
 apply_model_tier "economy"   # Set all consultants to economy models
 
 # Get model for a specific consultant and tier (v2.8.1)
-get_model_for_tier "gemini" "premium"   # → Gemini 3.1 Pro (High)
+get_model_for_tier "gemini" "premium"   # → Gemini 3.7 Flash (High) on agy
 get_model_for_tier "claude" "economy"   # → claude-haiku-4-5
 ```
 
@@ -392,26 +373,25 @@ get_model_for_tier "claude" "economy"   # → claude-haiku-4-5
 Three new presets leverage the model tiers:
 
 ```bash
-# Maximum quality - maximum tier + debate + peer review
+# Maximum quality - all consultants + maximum models and provider effort
 ./scripts/consult_all.sh --preset max_quality "critical architecture decision"
 
-# Balanced quality - standard models, 4 consultants, light debate
+# Balanced quality - standard models, 3 consultants
 ./scripts/consult_all.sh --preset medium "general coding question"
 
-# Super fast - economy models, 2 consultants, no debate
+# Super fast - economy models, 2 consultants
 ./scripts/consult_all.sh --preset fast "quick syntax question"
 ```
 
-### Premium Model Defaults (July 2026)
+### Premium Model Defaults (August 2026)
 All consultants now use premium models by default:
 
 | Consultant | Default Model |
 |------------|---------------|
 | Claude | claude-opus-5 |
-| Gemini | Gemini 3.1 Pro (High) (via agy CLI) |
+| Gemini | Gemini 3.7 Flash (High) (via agy CLI); API: gemini-3.1-pro-preview |
 | Codex | gpt-5.6-sol |
 | Mistral | CLI: mistral-medium-3.5; API: mistral-large-3 |
-| Cursor | composer-2.5 |
 | DeepSeek | deepseek-v4-pro |
 | GLM | glm-5.3 |
 | Grok | grok-4.6 |
@@ -430,6 +410,20 @@ missing, cannot launch, or lacks usable authentication and `GROK_API_KEY` is
 available. A timeout, model error, empty response, or other post-launch failure
 is surfaced and never silently rerouted to the API.
 
+The live full-panel smoke defines the `max_quality` runtime envelope, not just
+the model names. Mistral and Grok receive four bounded advisory turns plus an
+explicit no-tools/workspace instruction. Qwen3.8-Max gets a 600-second Token
+Plan timeout; DeepSeek V4 Pro/max gets 600 seconds and 16,384 completion tokens;
+GLM gets 16,384; MiniMax M3 gets 16,384 plus a compact Markdown contract in
+mmx's native system channel. OpenAI-compatible
+`finish_reason=length` is an error. Every
+response records `metadata.response_quality=structured|fallback|error`:
+malformed JSON-looking output fails closed, while real markdown/prose is kept
+as an explicit fallback. Synthesis filters error envelopes and includes bounded
+detail from structured and fallback responses. Claude CLI consultations are
+preflighted, stateless (`--no-session-persistence`), setting-source-free,
+tool-free, and MCP-free; a hung/help/auth probe fails before a paid dispatch.
+
 ## v2.4 Features
 
 ### Budget Enforcement (Opt-in)
@@ -449,8 +443,7 @@ BUDGET_ACTION=warn  # or "stop"
 **Enforcement Points:**
 1. Before Round 1 - Check estimated cost vs budget
 2. After Round 1 - Check actual cost vs warning threshold
-3. Before Debate - Check cumulative + debate estimate
-4. Before Synthesis - Check cumulative + synthesis estimate
+3. Before Synthesis - Check cumulative + synthesis estimate
 
 Functions in `lib/costs.sh`:
 - `is_budget_enabled()` - Check if budget enforcement is enabled
@@ -523,22 +516,6 @@ Functions in `lib/routing.sh`:
 - `get_premium_model()` - Returns premium model for consultant
 - `get_escalation_summary()` - Returns escalation info as JSON
 
-### Debate Optimization (Opt-in)
-Skips debate if confidence spread is low (all consultants agree).
-
-```bash
-ENABLE_DEBATE_OPTIMIZATION=false  # Default: false (opt-in per quality review)
-DEBATE_CONFIDENCE_SPREAD_THRESHOLD=3  # Min spread to trigger debate
-DEBATE_USE_SUMMARIES=true         # Use summaries in debate rounds
-```
-
-**Category Exceptions**: SECURITY and ARCHITECTURE always trigger debate regardless of confidence spread.
-
-Functions in `debate_round.sh`:
-- `is_mandatory_debate_category()` - Returns true for SECURITY/ARCHITECTURE
-- `should_skip_debate()` - Returns true if debate can be skipped
-- `extract_compact_summary()` - Token-optimized summary for debates
-
 ### Quality Monitoring
 Logs optimization metrics and saves to output directory.
 
@@ -555,12 +532,9 @@ Output file `optimization_metrics.json`:
     "cache_hits": 2,
     "response_limits_enabled": false,
     "cost_aware_routing": false,
-    "debate_optimization": false,
     "compact_report": true
   },
   "quality_metrics": {
-    "consensus_score": 75,
-    "consensus_level": "medium",
     "successful_responses": 4,
     "total_consultants": 4,
     "category": "ARCHITECTURE"
@@ -595,9 +569,6 @@ npm run lint
 # Test with preset
 ./scripts/consult_all.sh --preset minimal "Quick question"
 
-# Test with debate
-ENABLE_DEBATE=true ./scripts/consult_all.sh "Microservices or monolith?"
-
 # Test with strategy
 ./scripts/consult_all.sh --strategy risk_averse "Security question"
 
@@ -617,7 +588,7 @@ for f in scripts/*.sh scripts/lib/*.sh; do bash -n "$f" && echo "OK: $f"; done
 | `MAX_SESSION_COST` | 1.00 | Max budget ($) |
 | `CLAUDE_MODEL` | claude-opus-5 | Claude model (v2.5) |
 | `CLAUDE_API_MAX_TOKENS` | 16384 | Claude API thinking + visible-output budget |
-| `GEMINI_MODEL` | Gemini 3.1 Pro (High) | Gemini agy CLI model (v2.15) |
+| `GEMINI_MODEL` | Gemini 3.7 Flash (High) | Gemini agy CLI model |
 | `GEMINI_API_MODEL` | gemini-3.1-pro-preview | Gemini API-mode model ID (v2.15) |
 | `CODEX_MODEL` | gpt-5.6-sol | Codex model (v2.5) |
 | `MISTRAL_MODEL` | mistral-large-3 | Mistral model (v2.5) |
@@ -646,7 +617,6 @@ for f in scripts/*.sh scripts/lib/*.sh; do bash -n "$f" && echo "OK: $f"; done
 - `agy` CLI - Antigravity CLI (Gemini consultant; successor to the deprecated Gemini CLI, v2.15)
 - `codex` CLI - OpenAI Codex
 - `vibe` CLI - Mistral Vibe
-- `cursor-agent` CLI - Cursor (validated by capability, read-only ask mode)
 - `kimi` CLI - Kimi Code (v2.9)
 - `claude` CLI - Claude (v2.2, also used for synthesis)
 - `qwen` CLI - Qwen via qwen-code (v2.7, optional)
@@ -782,7 +752,7 @@ curl -fsSL https://raw.githubusercontent.com/matteoscurati/ai-consultants/main/s
 ### Guidelines
 
 - **Audience**: Developers who use the skill. Write for someone who hasn't seen the commits.
-- **Highlights first**: Lead with impact, not implementation. "Debate rounds no longer crash on Linux" > "Added `|| true` to `((count++))` in debate_round.sh".
+- **Highlights first**: Lead with impact, not implementation. "Provider failures are now diagnosed" > "Added an error branch".
 - **Breaking changes prominent**: If any exist, they go in a dedicated section — never buried in a bullet list.
 - **Quantify when possible**: Token savings percentages, issue counts, file counts.
 - **Upgrade guide always present**: Even if the answer is "just pull", make it explicit.
@@ -790,6 +760,25 @@ curl -fsSL https://raw.githubusercontent.com/matteoscurati/ai-consultants/main/s
 
 ## Changelog
 
+### v4.0.1
+
+- **The installer's first line of user-visible output was six major/minor generations stale.** `scripts/install.sh:53` printed `AI Consultants v2.10 - Installation` even while the public binary, package metadata, npm registry, active checkout, and doctor all reported v4.0.0. That did not change installed bytes or runtime behavior, but it made a successful v4 install look wrong at the exact moment users were deciding whether to trust it.
+- **The banner is now deliberately version-neutral, not dynamically coupled to a tenth version surface.** The curl-piped installer prints before it has cloned or updated the checkout, so sourcing `scripts/config.sh` there would either be impossible on a fresh install or would report the old local version during an update. A static `AI Consultants - Installation` label stays truthful throughout both paths and does not need another release-time rewrite.
+- **The failure is regression-tested through rendered output.** `scripts/test_install.sh` sources the existing define-only path, calls `print_header`, requires the installer label, and rejects `AI Consultants v<major>.<minor>` in the banner. The focused installer suite grows to 47 checks; the full 23-suite gate and ShellCheck pass on the 4.0.1 surfaces.
+
+  Deliberately unchanged: no runtime, model, preset, credential, host command, or minisite copy changes. The showcase site already reflects the ten-consultant v4 panel; after npm serves 4.0.1 it needs only its two mechanical version surfaces bumped.
+
+### v4.0.0
+
+- **`max_quality` now means the full ten-consultant panel at each exact transport's strongest supported effort and response budget.** GLM and DeepSeek run at `max`; Grok uses its highest supported setting. K3-256k, MiniMax M3, and Qwen3.8-Max remain confined to the maximum tier, with Qwen promoted only when the authenticated Token Plan transport is already configured.
+- **Gemini 3.7 Flash High completed an authenticated exact-adapter `agy` smoke and is promoted only on that CLI transport.** It serves maximum, premium, and standard; Low serves economy. The unverified Google API transport remains on 3.1 Pro, while Fable 5 and the new Mistral API IDs remain opt-in.
+- **Cursor is no longer a consultant.** The ambiguous `agent`/`cursor-agent` runtime, model tiers, presets, doctor/configuration fields, personas, costs, tests, and website card were removed. Cursor remains a supported host through SkillPort; the panel itself has 10 consultants.
+- **Managed migrations are exact and pins survive.** `configure` upgrades only the historical managed Gemini 3.1 Pro CLI, GLM 5.2, and Grok 4.5 values, while removing obsolete Cursor settings. Environment/`--set` model values and any `# ai-consultants:pin` entry remain user-owned. The catalog-parity gate checks all 40 consultant/tier cells and all 16 Gemini/Mistral transport cells against `docs/cost_rates.json`, requiring every automatic target to be priced or explicitly unpriced.
+- **Claude can launch the skill as a real host, not merely describe it.** `SKILL.md` carries an imperative execution contract; all host commands use a private query-file handoff, preserve preset/strategy options, pass the caller project through `--context-root`, and set `INVOKING_AGENT`; self-exclusion covers consultation, full-roster targeted follow-ups, and synthesis for Claude/Codex/Gemini. Offline E2E proves a Claude host dispatches Gemini/Codex/Mistral with no Claude artifacts. A real Claude Code/Sonnet run started from an external temp project, used `mktemp`, mode 600, a quoted heredoc, trap cleanup, `--context-root`, and `src/context.txt@CONTEXT`; Gemini 3.7 High and Codex 5.6 Sol returned (2/2), the external marker/tag reached both, no Claude artifact appeared, and the private staging directory was removed. The live run disabled synthesis; symmetric regressions prove each host is excluded there. Current runtime prompts, schemas, templates, help, and instructions are coverage-only; stale Codex/Gemini debate and roster-audit commands are removed.
+- **Cross-family host review went `NO_SHIP` → `NO_SHIP` → `SHIP`.** The review found the Bash 3.2 empty-array abort, query-file/context parser conflict, caller-project path loss, unsafe/host-self follow-ups, stale live consensus output, broad uninstall glob, and active v3.0 residue. Each was fixed with executable coverage. After the `SHIP` verdict, its remaining low-severity findings were also resolved: staged files retain their original project-relative identity, fallback synthesis reports the requested strategy, follow-up context no longer rides in argv and respects disabled consultants, and remaining contributor/cost/test residue was removed.
+
+  Deliberate residuals: Vibe isolates the workspace but retains ambient HOME for its authenticated CLI state; unpriced credit/subscription models are disclosed but cannot be enforced by a dollar budget; provider/requested model substitutions are available in raw JSON but not yet summarized in the human report.
+
 ### v3.5.0
 
 - **The catalog now distinguishes an advisory panel's normal premium tier from subscription-only maximum targets.** `get_model_for_tier()` accepts an explicit transport, while `apply_model_tier maximum` selects K3-256k and MiniMax M3 and promotes Qwen3.8-Max only when `QWEN3_USE_API=true`, the wire is OpenAI-compatible, the endpoint is a Token Plan `/chat/completions` URL, and a key is present (`scripts/config.sh:581-737`). The Qwen effort is marked tier-managed, cleared when leaving maximum, and never overwrites a user pin. Gemini 3.7, Fable 5, and the new Mistral API IDs remain catalogued opt-ins after their exact transports failed authentication or timed out; normal runs do not inherit delegation-kit's routing gates.