ai-consultants 2.10.1

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

@@ -0,0 +1,94 @@
+---
+description: Consult AI experts (Gemini, Codex, Mistral, Kilo, Cursor, Aider, Amp, Kimi, Qwen, GLM, Grok, DeepSeek, MiniMax, Ollama) 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
+---
+
+# AI Consultants - Expert Panel
+
+Query multiple AI models for expert opinions on a coding question.
+
+**User's Question:** $ARGUMENTS
+
+## Instructions
+
+1. If no question provided in $ARGUMENTS, ask the user what they want to consult about.
+
+2. **File context handling**: Check if $ARGUMENTS contains file paths (tokens containing `/` or extensions like `.ts`, `.js`, `.py`, `.sh`, `.json`, `.yaml`, `.yml`, `.md`, `.go`, `.rs`, `.java`, `.rb`, `.css`, `.html`, `.tsx`, `.jsx`, `.c`, `.cpp`, `.h`).
+   If file paths are found:
+   - Extract each file path token from $ARGUMENTS
+   - Read each file using the **Read** tool
+   - Build an enriched query by replacing file path tokens with their inline contents:
+     ```
+     <original question text without file paths>
+
+     --- File: path/to/file.ts ---
+     <file contents>
+     --- End File ---
+     ```
+   - Use this enriched query in step 3 instead of raw $ARGUMENTS
+
+3. **Run the consultation**:
+   ```bash
+   cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && INVOKING_AGENT=claude ./scripts/consult_all.sh '<query>'
+   ```
+   Replace `<query>` with the enriched query from step 2 (or $ARGUMENTS if no files were detected). Use single quotes around the query to prevent shell expansion of special characters.
+   **Capture the last line of stdout** — it contains the output directory path (e.g., `/tmp/ai_consultations/20260315_143022_12345/`).
+
+4. **Read the results** from the output directory captured in step 3:
+   - Read `<output_dir>/report.md` using the **Read** tool
+   - Read `<output_dir>/synthesis.json` using the **Read** tool (may not exist if synthesis was skipped or failed)
+   - If neither exists, use the **Glob** tool with pattern `<output_dir>/*.json` to discover individual consultant response files, then read each one
+
+5. **Present the results** using this template:
+
+   ### Consensus
+   - **Score**: <consensus_percentage>% (<consensus_level: strong/medium/low/none>)
+   - **Category**: <question_category>
+
+   ### Consultant Summary
+   | Consultant | Confidence | Approach | Key Insight |
+   |------------|-----------|----------|-------------|
+   | <name> | <score>/10 | <approach> | <one-line from summary> |
+
+   ### Synthesized Recommendation
+   - **Recommended approach**: <approach>
+   - **Summary**: <synthesized summary>
+
+   ### Key Agreements
+   - <points most consultants agree on>
+
+   ### Key Disagreements
+   - <points where consultants diverge>
+
+   ### Risk Assessment
+   - <caveats and uncertainty factors raised>
+
+   ### Suggested Next Steps
+   - <actionable items based on the recommendation>
+
+   If `synthesis.json` is missing, build the summary table and recommendation directly from the individual consultant JSON files in the output directory.
+
+6. **Error recovery**:
+   - **Exit code 1** (general failure): Tell the user and suggest running diagnostics:
+     ```bash
+     cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && ./scripts/doctor.sh
+     ```
+   - **Exit code 124** (timeout): Suggest `--preset fast` for quicker results or increasing the timeout (e.g., `GEMINI_TIMEOUT=300`)
+   - **Partial failures** (some consultants responded, others didn't): Present the successful results normally and note which consultants failed at the end
+   - **Empty output directory** (no JSON files produced): Suggest running `doctor.sh` to verify CLI installations and API keys
+
+## Options
+
+| Variable | Effect |
+|----------|--------|
+| `ENABLE_DEBATE=true` | Enable multi-round debate for deeper analysis |
+| `ENABLE_SMART_ROUTING=true` | Auto-select best consultants for the question type |
+| `DEBATE_ROUNDS=2` | Number of debate rounds (1-3) |
+
+## Follow-up
+
+For follow-up questions on the same consultation:
+```bash
+cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && INVOKING_AGENT=claude ./scripts/followup.sh '<follow-up question>'
+```

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

@@ -0,0 +1,103 @@
+---
+description: Run AI consultation with multi-round debate where consultants critique each other's positions. Use for architecture decisions, security reviews, controversial trade-offs, or when a standard consultation produced low consensus.
+argument-hint: <question> [file1] [file2] ...
+allowed-tools: Bash Read Glob Grep
+---
+
+# AI Consultants - Debate Mode
+
+Run a multi-round debate where AI consultants critique each other's responses and refine their positions.
+
+**Question:** $ARGUMENTS
+
+## When to Use Debate
+
+- **Architecture decisions**: Microservices vs monolith, database choices, system design
+- **Security reviews**: Threat modeling, authentication strategies, data protection approaches
+- **Low-consensus results**: A previous `/ai-consultants:consult` returned low consensus — debate forces deeper analysis
+- **Controversial trade-offs**: Performance vs maintainability, speed vs correctness, build vs buy
+
+## Instructions
+
+1. If no question provided in $ARGUMENTS, ask the user what they want to debate.
+
+2. **File context handling**: Check if $ARGUMENTS contains file paths (tokens containing `/` or extensions like `.ts`, `.js`, `.py`, `.sh`, `.json`, `.yaml`, `.yml`, `.md`, `.go`, `.rs`, `.java`, `.rb`, `.css`, `.html`, `.tsx`, `.jsx`, `.c`, `.cpp`, `.h`).
+   If file paths are found:
+   - Extract each file path token from $ARGUMENTS
+   - Read each file using the **Read** tool
+   - Build an enriched query by replacing file path tokens with their inline contents:
+     ```
+     <original question text without file paths>
+
+     --- File: path/to/file.ts ---
+     <file contents>
+     --- End File ---
+     ```
+   - Use this enriched query in step 3 instead of raw $ARGUMENTS
+
+3. **Run the debate consultation**:
+   ```bash
+   cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && INVOKING_AGENT=claude ENABLE_DEBATE=true DEBATE_ROUNDS="${DEBATE_ROUNDS:-2}" ./scripts/consult_all.sh '<query>'
+   ```
+   Replace `<query>` with the enriched query from step 2 (or $ARGUMENTS if no files were detected). Use single quotes around the query to prevent shell expansion of special characters.
+   **Capture the last line of stdout** — it contains the output directory path.
+
+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)
+   - Use the **Glob** tool with pattern `<output_dir>/round_*/round_summary.json` to find debate round summaries, then read each one
+   - Use the **Glob** tool with pattern `<output_dir>/round_*/*.json` to find debate round responses (exclude `round_summary.json` — these individual files contain position changes, critiques, and confidence deltas)
+   - Round 1 (initial responses) are stored directly in `<output_dir>/*.json`; subsequent rounds are in `<output_dir>/round_N/` subdirectories
+
+5. **Present the results** using this debate-specific template:
+
+   ### Consensus
+   - **Score**: <consensus_percentage>% (<consensus_level>)
+   - **Category**: <question_category>
+
+   ### Position Evolution
+   | Consultant | Initial Approach | Changed? | Final Approach | Confidence Delta |
+   |------------|-----------------|----------|---------------|-----------------|
+   | <name> | <round 1 approach> | Yes/No | <final approach> | <+N/-N> |
+
+   ### Key Critiques
+   - **<consultant>** on **<target>**: <critique summary> (severity: <minor/moderate/major>)
+
+   ### Debate Stability
+   - **Assessment**: <stable/mostly_stable/volatile> — <N> of <total> consultants changed position
+   - If stable: "Strong agreement — consultants held their positions after cross-critique"
+   - If volatile: "Significant disagreement — positions shifted substantially during debate"
+
+   ### Synthesized Recommendation
+   - **Recommended approach**: <approach>
+   - **Summary**: <synthesized summary>
+
+   ### Key Agreements
+   - <points reinforced through debate>
+
+   ### Key Disagreements
+   - <unresolved points of contention>
+
+   ### Risk Assessment
+   - <caveats and uncertainty factors>
+
+   ### Suggested Next Steps
+   - <actionable items>
+
+   If `synthesis.json` is missing, build the summary from individual consultant JSON files.
+
+6. **Error recovery**:
+   - **Exit code 1** (general failure): Tell the user and suggest running diagnostics:
+     ```bash
+     cd "${AI_CONSULTANTS_DIR:-$HOME/.claude/skills/ai-consultants}" && ./scripts/doctor.sh
+     ```
+   - **Exit code 124** (timeout): Debate is more expensive — suggest reducing `DEBATE_ROUNDS=1` or using `--preset fast`
+   - **Partial failures**: Present successful results and note which consultants failed
+   - **Empty output**: Suggest running `doctor.sh` to verify CLI installations
+
+## Options
+
+| Variable | Default | Effect |
+|----------|---------|--------|
+| `DEBATE_ROUNDS` | `2` | Number of debate rounds (1-3). More rounds = deeper analysis but higher cost |
+| `ENABLE_SMART_ROUTING` | `false` | Auto-select best consultants for the question type |

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

@@ -0,0 +1,30 @@
+---
+description: Show help and usage for AI Consultants skill
+allowed-tools:
+---
+
+# AI Consultants - Help
+
+All skill details (consultants, personas, presets, strategies, features, configuration) are shown in the skill overview loaded alongside this command. Here's a quick reference:
+
+## Commands
+
+| Command | Description |
+|---------|-------------|
+| `/ai-consultants:consult` | Main consultation — ask AI consultants a coding question |
+| `/ai-consultants:debate` | Run consultation with multi-round debate |
+| `/ai-consultants:help` | Show this help |
+
+## Quick Examples
+
+```
+/ai-consultants:consult "How to optimize this SQL query?"
+/ai-consultants:consult "Review this code" src/utils.ts
+/ai-consultants:debate "Microservices or monolith?"
+```
+
+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`.
+
+Full docs: https://github.com/matteoscurati/ai-consultants