multi-model-meta-analysis
Synthesize outputs from multiple AI models into a comprehensive, verified assessment. Use when: (1) User pastes feedback/analysis from multiple LLMs (Claude, GPT, Gemini, etc.) about code or a project, (2) User wants to consolidate model outputs into a single reliable document, (3) User needs conflicting model claims resolved against actual source code. This skill verifies model claims against the codebase, resolves contradictions with evidence, and produces a more reliable assessment than any single model.
What this skill does
# Multi-Model Synthesis
Combine outputs from multiple AI models into a verified, comprehensive assessment by cross-referencing claims against the actual codebase.
## Core Principle
Models hallucinate and contradict each other. The source code is the source of truth. Every significant claim must be verified before inclusion in the final assessment.
## Process
### 1. Extract Claims
Parse each model's output and extract discrete claims:
- Factual assertions about the code ("function X does Y", "there's no error handling in Z")
- Recommendations ("should add validation", "refactor this pattern")
- Identified issues ("bug in line N", "security vulnerability")
Tag each claim with its source model.
### 2. Deduplicate
Group semantically equivalent claims:
- "Lacks input validation" = "No sanitization" = "User input not checked"
- "Should use async/await" = "Convert to promises" = "Make asynchronous"
Create canonical phrasing. Track which models mentioned each.
### 3. Verify Against Source
For each factual claim or identified issue:
```
CLAIM: "The auth middleware doesn't check token expiry"
VERIFY: Read the auth middleware file
FINDING: [Confirmed | Refuted | Partially true | Cannot verify]
EVIDENCE: [Quote relevant code or explain why claim is wrong]
```
Use Grep, Glob, and Read tools to locate and examine relevant code. Do not trust model claims without verification.
### 4. Resolve Conflicts
When models contradict each other:
1. Identify the specific disagreement
2. Examine the actual code
3. Determine which model (if any) is correct
4. Document the resolution with evidence
```
CONFLICT: Model A says "uses SHA-256", Model B says "uses MD5"
INVESTIGATION: Read crypto.js lines 45-60
RESOLUTION: Model B is correct - line 52 shows MD5 usage
EVIDENCE: `const hash = crypto.createHash('md5')`
```
### 5. Synthesize Assessment
Produce a final document that:
- States verified facts (not model opinions)
- Cites evidence for significant claims
- Notes where verification wasn't possible
- Preserves valuable insights that don't require verification (e.g., design suggestions)
## Output Format
```markdown
# Synthesized Assessment: [Topic]
## Summary
[2-3 sentences describing the verified findings]
## Verified Findings
### Confirmed Issues
| Issue | Severity | Evidence | Models |
|-------|----------|----------|--------|
| [Issue] | High/Med/Low | [file:line or quote] | Claude, GPT |
### Refuted Claims
| Claim | Source | Reality |
|-------|--------|---------|
| [What model said] | GPT-4 | [What code actually shows] |
### Unverifiable Claims
| Claim | Source | Why Unverifiable |
|-------|--------|------------------|
| [Claim] | Claude | [Requires runtime testing / external system / etc.] |
## Consensus Recommendations
[Items where 2+ models agree AND verification supports the suggestion]
## Unique Insights Worth Considering
[Valuable suggestions from single models that weren't contradicted]
## Conflicts Resolved
| Topic | Model A | Model B | Verdict | Evidence |
|-------|---------|---------|---------|----------|
| [Topic] | [Position] | [Position] | [Which is correct] | [Code reference] |
## Action Items
### Critical (Verified, High Impact)
- [ ] [Item] — Evidence: [file:line]
### Important (Verified, Medium Impact)
- [ ] [Item] — Evidence: [file:line]
### Suggested (Unverified but Reasonable)
- [ ] [Item] — Source: [Models]
```
## Verification Guidelines
**Always verify:**
- Bug reports and security issues
- Claims about what code does or doesn't do
- Assertions about missing functionality
- Performance or complexity claims
**Trust but note source:**
- Style and readability suggestions
- Architectural recommendations
- Best practice suggestions
**Mark as unverifiable:**
- Runtime behavior claims (without tests)
- Performance benchmarks (without profiling)
- External API behavior
- User experience claims
## Anti-Patterns
- Blindly merging model outputs without checking code
- Treating model consensus as proof (all models can be wrong)
- Omitting refuted claims (document what was wrong - it's valuable)
- Skipping verification because claims "sound right"
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.