{ "schema_version": "https://schema.org", "type": "DataFeed", "name": "AKI AI Vibe Coding & Debugging Intelligence Atlas Data Feed", "description": "Actuarial developer intelligence evaluating 34 global vibe coding ecosystems, debugging burden scores, reality gaps, 7 avoidance rules, and decision matrix.", "canonical_url": "https://aki1k.com/vibecoding", "snapshot_date": "2026-10-08", "edition": "v3.2 \"Anti-Hype Grounded\"", "governance": "ZIP-1.0 Actuarial Governance", "macro": { "edition": "v3.2 \"Anti-Hype Grounded\"", "snapshot_date": "2026-10-08", "governance": "ZIP-1.0 Actuarial Governance", "total_ecosystems": 34, "lowest_burden_tool": "Claude Code (24.2 pts)", "lowest_reality_gap_tool": "Aider (+7.8 pts)", "highest_moat_tool": "GitHub Copilot (94.2 pts)", "highest_autonomy_tool": "Devin (93.2%)", "peak_aki_score": 91.7, "peak_aki_tool": "Google Antigravity 2.0", "avg_reality_gap_delta": 12.6, "signature_principle": "More generated code does not necessarily mean more productive software engineering.", "the_70_percent_wall": { "zone_1": "0–70%: AI accelerates development — rapid prototypes, instant boilerplate", "zone_2": "70–85%: Debugging exceeds creation time; context loss leads to circular fix loops", "zone_3": "85–100%: Regression rate triples; generated changes break existing production contracts", "burden_tiers": { "sustainable": "20–30 pts: Sustainable production — low rework, disciplined commits", "manageable": "30–40 pts: Good with tests — strict human diff review required", "debt_heavy": "40+ pts: Prototype-only — high long-term technical debt" } }, "anti_hype_guards": [ "exclude_tiktok_views", "exclude_arr_and_valuation", "exclude_github_stars_from_ranking", "cap_vendor_claim_at_50pct", "discount_self_reported_autonomy_40pct" ] }, "refresh_spec": { "version": "3.2", "governance": "ZIP-1.0", "scoring": "popularity_weight_0", "refresh_cron": "0 3 * * *", "sources": [ "https://aki1k.com/api/ecosystems.json", "https://benchlm.ai/vibe-code-bench", "https://github.com/awesome-vibecoding", "https://codescene.com/research", "https://lightrun.com/state-of-ai-engineering-2026" ], "anti_hype_guards": [ "exclude_tiktok_views", "exclude_arr", "exclude_github_stars_from_score", "cap_vendor_claim_50pct", "discount_self_reported_autonomy_40pct" ] }, "ecosystems_count": 33, "ecosystems": [ { "id": "vc_aider", "name": "Aider", "slug": "aider", "category": "A", "category_rank": "A1", "category_name": "Terminal / CLI Agents", "ecosystem_type": "terminal_agent", "developer_org": "OSS / Paul Gauthier", "primary_strength": "Git-native terminal pairing, architect/editor dual-model loop, local commit gating", "maturity_tier": 1, "performance_score": 90.1, "moat_number": 84.5, "debugging_burden_score": 29.8, "reality_gap_points": 7.8, "production_reliability": 85.1, "agentic_autonomy": 81.2, "evidence_state": "AKI Verified", "confidence": 0.95, "methodology_version": "v3.2.0", "badge": "🏆 #1 Lowest Reality Gap (+7.8)", "notes": "Lowest gap; Git-native; proven merge history in peer-reviewed studies.", "defect_breakdown": { "syntax_pct": 10.5, "logic_pct": 35, "context_drift_pct": 22.5, "hallucinated_imports_pct": 32 }, "mttr_mins": 11.5, "rule_snippet": "# .aider.conf.yml\nauto-commits: false\nlint-cmd: \"npm run lint\"\ntest-cmd: \"npm test\"\nattribute-author: true", "cli_command": "aider --model anthropic/claude-3-7-sonnet --test-cmd \"npm test\"" }, { "id": "vc_claude_code", "name": "Claude Code", "slug": "claude-code", "category": "A", "category_rank": "A2", "category_name": "Terminal / CLI Agents", "ecosystem_type": "terminal_agent", "developer_org": "Anthropic", "primary_strength": "Long-horizon terminal coding, multi-file reasoning, native Bash agent sandbox", "maturity_tier": 1, "performance_score": 89.2, "moat_number": 89, "debugging_burden_score": 24.2, "reality_gap_points": 9.7, "production_reliability": 89.2, "agentic_autonomy": 91, "evidence_state": "AKI Verified", "confidence": 0.95, "methodology_version": "v3.2.0", "badge": "Lowest Repair Burden (24.2)", "notes": "Strongest reliability; reality gap corrected from +8.4 to +9.7 post-merge.", "defect_breakdown": { "syntax_pct": 8, "logic_pct": 31, "context_drift_pct": 28, "hallucinated_imports_pct": 33 }, "mttr_mins": 8.5, "rule_snippet": "# CLAUDE.md\n- Enforce strict TypeScript: no \"any\"\n- Always run local tests before returning control\n- Scope edits to <= 3 files per tool pass", "cli_command": "claude --dangerously-skip-permissions false" }, { "id": "vc_codex_cli", "name": "OpenAI Codex CLI", "slug": "codex-cli", "category": "A", "category_rank": "A3", "category_name": "Terminal / CLI Agents", "ecosystem_type": "terminal_agent", "developer_org": "OpenAI", "primary_strength": "Terminal-Bench leader, low-level shell tool invocation, deterministic bash planning", "maturity_tier": 2, "performance_score": 88.7, "moat_number": 83, "debugging_burden_score": 28.4, "reality_gap_points": 9.1, "production_reliability": 84.8, "agentic_autonomy": 86.5, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Terminal-Bench Leader", "notes": "Consistent terminal evaluation results; highly predictable agent trajectory.", "defect_breakdown": { "syntax_pct": 9.5, "logic_pct": 34, "context_drift_pct": 26.5, "hallucinated_imports_pct": 30 }, "mttr_mins": 10.2, "rule_snippet": "# CODEX.md\n- Verify command syntax with --dry-run where applicable\n- Keep bash subshells isolated", "cli_command": "codex run --approval-mode manual" }, { "id": "vc_qwen_code", "name": "Qwen Code + Qwen3-Coder", "slug": "qwen-code", "category": "A", "category_rank": "A4", "category_name": "Terminal / CLI Agents", "ecosystem_type": "open_agent", "developer_org": "Alibaba Cloud", "primary_strength": "Open-weight state-of-the-art coding, terminal agent adaptability, extreme inference value", "maturity_tier": 2, "performance_score": 88.5, "moat_number": 82.5, "debugging_burden_score": 31.2, "reality_gap_points": 9.9, "production_reliability": 83.7, "agentic_autonomy": 85, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Open-Weight Cost Leader", "notes": "Fastest-improving open stack; within 3% of top proprietary on SWE-bench.", "defect_breakdown": { "syntax_pct": 12, "logic_pct": 36, "context_drift_pct": 24, "hallucinated_imports_pct": 28 }, "mttr_mins": 12.8, "rule_snippet": "# qwen_agent.json\n{\n \"max_context_window\": 131072,\n \"enforce_strict_cot\": true\n}", "cli_command": "qwen-code --local-model qwen3-coder-32b" }, { "id": "vc_gemini_cli", "name": "Gemini CLI", "slug": "gemini-cli", "category": "A", "category_rank": "A5", "category_name": "Terminal / CLI Agents", "ecosystem_type": "terminal_agent", "developer_org": "Google", "primary_strength": "Direct Google Antigravity cloud runtime connection, 2M context ingestion, zero setup", "maturity_tier": 2, "performance_score": 86.4, "moat_number": 81, "debugging_burden_score": 30.1, "reality_gap_points": 9.5, "production_reliability": 83.2, "agentic_autonomy": 82, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "2M Context Terminal", "notes": "Antigravity integration; generous free tier drives massive global adoption.", "defect_breakdown": { "syntax_pct": 11, "logic_pct": 37, "context_drift_pct": 22, "hallucinated_imports_pct": 30 }, "mttr_mins": 12, "rule_snippet": "# .geminirc\ncontext_window: 2000000\ncode_execution: sandbox", "cli_command": "gemini code --sandbox true" }, { "id": "vc_kimi_code", "name": "Kimi Code", "slug": "kimi-code", "category": "A", "category_rank": "A6", "category_name": "Terminal / CLI Agents", "ecosystem_type": "terminal_agent", "developer_org": "Moonshot AI", "primary_strength": "Exceptional ultra-long context preservation, full repository dependency tree mapping", "maturity_tier": 2, "performance_score": 85.3, "moat_number": 78.5, "debugging_burden_score": 30.8, "reality_gap_points": 9.8, "production_reliability": 82.9, "agentic_autonomy": 83.5, "evidence_state": "AKI Verified", "confidence": 0.93, "methodology_version": "v3.2.0", "badge": "Long-Context Retention", "notes": "Superior long-context retention in multi-hour coding sessions.", "defect_breakdown": { "syntax_pct": 13, "logic_pct": 38, "context_drift_pct": 21, "hallucinated_imports_pct": 28 }, "mttr_mins": 13.5, "rule_snippet": "# kimi.yaml\ncontext_cache: true\ntrace_dependencies: true", "cli_command": "kimi-cli --repo-mode" }, { "id": "vc_deepseek_harness", "name": "DeepSeek Harness", "slug": "deepseek-harness", "category": "A", "category_rank": "A7", "category_name": "Terminal / CLI Agents", "ecosystem_type": "open_agent", "developer_org": "DeepSeek", "primary_strength": "Unmatched reasoning-per-dollar, deep architectural planning loops, self-correction", "maturity_tier": 2, "performance_score": 83.8, "moat_number": 79, "debugging_burden_score": 32.1, "reality_gap_points": 10.3, "production_reliability": 81.1, "agentic_autonomy": 83, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Reasoning-per-$ SOTA", "notes": "Best reasoning-per-dollar; overtaking premium commercial tiers in value.", "defect_breakdown": { "syntax_pct": 14, "logic_pct": 35, "context_drift_pct": 25, "hallucinated_imports_pct": 26 }, "mttr_mins": 14.2, "rule_snippet": "# deepseek_rules.md\n- Enforce chain-of-thought verification on intermediate AST nodes", "cli_command": "deepseek-harness --reasoning-effort high" }, { "id": "vc_openhands", "name": "OpenHands", "slug": "openhands", "category": "A", "category_rank": "A8", "category_name": "Terminal / CLI Agents", "ecosystem_type": "open_agent", "developer_org": "OSS Community", "primary_strength": "Autonomous open-source software engineering agent, containerized runtime sandbox", "maturity_tier": 2, "performance_score": 82.4, "moat_number": 76, "debugging_burden_score": 33.5, "reality_gap_points": 10.2, "production_reliability": 80.4, "agentic_autonomy": 84.2, "evidence_state": "AKI Verified", "confidence": 0.93, "methodology_version": "v3.2.0", "badge": "100% Open Community", "notes": "Fully open; active community development; transparent docker runtime.", "defect_breakdown": { "syntax_pct": 15, "logic_pct": 36, "context_drift_pct": 26, "hallucinated_imports_pct": 23 }, "mttr_mins": 15, "rule_snippet": "# config.toml\n[security]\nrequire_confirmation = true\nsandbox_timeout = 600", "cli_command": "openhands --headless --task \"Resolve issue #42\"" }, { "id": "vc_antigravity", "name": "Google Antigravity 2.0", "slug": "google-antigravity", "category": "B", "category_rank": "B1", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "orchestrator", "developer_org": "Google", "primary_strength": "Agent-first cloud workspace, automated CLI/SDK multi-agent synthesis, Hermes coordination", "maturity_tier": 1, "performance_score": 91.7, "moat_number": 92, "debugging_burden_score": 26, "reality_gap_points": 10.1, "production_reliability": 88, "agentic_autonomy": 89.5, "evidence_state": "AKI Verified", "confidence": 0.95, "methodology_version": "v3.2.0", "badge": "🏆 #1 Highest Moat (92.0)", "notes": "Highest moat; multi-agent reliability up 12% in Q3-Q4 2026 logs.", "defect_breakdown": { "syntax_pct": 7.5, "logic_pct": 32, "context_drift_pct": 28.5, "hallucinated_imports_pct": 32 }, "mttr_mins": 9.8, "rule_snippet": "# .agents/rules/AGENTS.md\n- Enforce Simulation Phase before modifying files\n- Cap subagent nesting depth at 2", "cli_command": "agy task start --profile autonomous" }, { "id": "vc_cursor", "name": "Cursor", "slug": "cursor", "category": "B", "category_rank": "B2", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "ai_ide", "developer_org": "Anysphere", "primary_strength": "IDE native context engine, Composer multi-agent workflows, tab predictive autocomplete", "maturity_tier": 1, "performance_score": 90.8, "moat_number": 86.5, "debugging_burden_score": 28.6, "reality_gap_points": 9.2, "production_reliability": 86.4, "agentic_autonomy": 84.5, "evidence_state": "AKI Verified", "confidence": 0.95, "methodology_version": "v3.2.0", "badge": "🏆 #1 AI IDE Daily Driver", "notes": "Top workflow ergonomics; ARR and valuation strictly excluded from score.", "defect_breakdown": { "syntax_pct": 10, "logic_pct": 33, "context_drift_pct": 27, "hallucinated_imports_pct": 30 }, "mttr_mins": 11.2, "rule_snippet": "# .cursorrules\n- You are pairing with an engineer; prefer minimal diffs\n- Never overwrite entire files without explicit confirmation", "cli_command": "cursor . --composer" }, { "id": "vc_kiro", "name": "Kiro", "slug": "kiro", "category": "B", "category_rank": "B3", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "orchestrator", "developer_org": "AWS / Independent", "primary_strength": "Specification-driven formal engineering and automated verification loops", "maturity_tier": 2, "performance_score": 87.5, "moat_number": 75, "debugging_burden_score": 27.4, "reality_gap_points": 9.8, "production_reliability": 86, "agentic_autonomy": 80.5, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Spec-First Rigor", "notes": "Spec-first architecture guarantees lowest architectural drift across releases.", "defect_breakdown": { "syntax_pct": 8.5, "logic_pct": 30, "context_drift_pct": 29, "hallucinated_imports_pct": 32.5 }, "mttr_mins": 10.5, "rule_snippet": "# .kiro/spec.json\n{\n \"require_formal_contract\": true,\n \"enforce_invariant_checks\": true\n}", "cli_command": "kiro verify --strict" }, { "id": "vc_cline_roo", "name": "Cline / Roo Code", "slug": "cline-roo-code", "category": "B", "category_rank": "B4", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "ai_ide", "developer_org": "Open Source", "primary_strength": "Autonomous coding in VS Code with human-in-the-loop tool approvals & role switching", "maturity_tier": 2, "performance_score": 86.4, "moat_number": 74, "debugging_burden_score": 34, "reality_gap_points": 11, "production_reliability": 80.8, "agentic_autonomy": 82, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Best HITL Controls", "notes": "Best human-in-the-loop controls; escape rate reduced through approval prompts.", "defect_breakdown": { "syntax_pct": 12, "logic_pct": 35, "context_drift_pct": 26, "hallucinated_imports_pct": 27 }, "mttr_mins": 14.5, "rule_snippet": "# .clinerules\n- Always ask before editing database migration files\n- Run npm test after every file write", "cli_command": "code --install-extension roovet.roo-cline" }, { "id": "vc_muse_agent", "name": "Muse Agent / Muse Code", "slug": "muse-agent", "category": "B", "category_rank": "B5", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "open_agent", "developer_org": "Meta", "primary_strength": "Fast-rising open-weight multimodal coding engine, multimodal AST inspection", "maturity_tier": 2, "performance_score": 85.9, "moat_number": 78, "debugging_burden_score": 29.5, "reality_gap_points": 10.4, "production_reliability": 83.5, "agentic_autonomy": 84, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Oct 2026 Open Boost", "notes": "Fast-rising open-weight stack; October 2026 independent evidence boost.", "defect_breakdown": { "syntax_pct": 11.5, "logic_pct": 34.5, "context_drift_pct": 26, "hallucinated_imports_pct": 28 }, "mttr_mins": 12.2, "rule_snippet": "# muse_rules.yaml\nmode: code_architect\nverify_ast_clean: true", "cli_command": "muse-code --serve-local --gpu-auto" }, { "id": "vc_trae", "name": "Trae", "slug": "trae", "category": "B", "category_rank": "B6", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "ai_ide", "developer_org": "ByteDance", "primary_strength": "Adaptive AI-native IDE with collaborative agent-driven builder workspace", "maturity_tier": 3, "performance_score": 85.7, "moat_number": 73, "debugging_burden_score": 37.5, "reality_gap_points": 10.8, "production_reliability": 79, "agentic_autonomy": 78, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "ByteDance Workspace", "notes": "Strong APAC enterprise adoption; context handling improving rapidly.", "defect_breakdown": { "syntax_pct": 13, "logic_pct": 37, "context_drift_pct": 25, "hallucinated_imports_pct": 25 }, "mttr_mins": 15.2, "rule_snippet": "# trae.config\nagent_collaborator: true\nauto_lint: true", "cli_command": "trae ." }, { "id": "vc_qoder_cn", "name": "Qoder CN", "slug": "qoder-cn", "category": "B", "category_rank": "B7", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "ai_ide", "developer_org": "Alibaba", "primary_strength": "Enterprise engineering copilot, deep repo-level reasoning, successor to Lingma", "maturity_tier": 3, "performance_score": 85.8, "moat_number": 80, "debugging_burden_score": 38.2, "reality_gap_points": 11.2, "production_reliability": 80.4, "agentic_autonomy": 75.5, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "China Enterprise #1", "notes": "China enterprise engineering leader; excellent multilingual code synthesis.", "defect_breakdown": { "syntax_pct": 12.5, "logic_pct": 36.5, "context_drift_pct": 26, "hallucinated_imports_pct": 25 }, "mttr_mins": 15.6, "rule_snippet": "# .qoderrules\nsecurity_check: strict\ncompliance_gate: enterprise", "cli_command": "qoder start" }, { "id": "vc_copilot", "name": "GitHub Copilot", "slug": "github-copilot", "category": "B", "category_rank": "B8", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "ai_ide", "developer_org": "Microsoft / GitHub", "primary_strength": "Deep ecosystem distribution, enterprise workspace context, pull request agents", "maturity_tier": 3, "performance_score": 85.2, "moat_number": 94.2, "debugging_burden_score": 39.4, "reality_gap_points": 14.8, "production_reliability": 81.5, "agentic_autonomy": 72, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Enterprise Distribution #1", "notes": "Distribution does not equal performance; reality gap is substantial.", "defect_breakdown": { "syntax_pct": 14, "logic_pct": 38, "context_drift_pct": 24, "hallucinated_imports_pct": 24 }, "mttr_mins": 16.8, "rule_snippet": "# .github/copilot-instructions.md\n- Limit suggested code changes to selected method signature", "cli_command": "gh copilot suggest \"Explain PR diff\"" }, { "id": "vc_windsurf", "name": "Windsurf / Cascade", "slug": "windsurf", "category": "B", "category_rank": "B9", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "ai_ide", "developer_org": "Codeium", "primary_strength": "Cascade flow-state editing, deep indexing across entire codebase, variable tracking", "maturity_tier": 2, "performance_score": 84.9, "moat_number": 77, "debugging_burden_score": 31.5, "reality_gap_points": 10.5, "production_reliability": 83.5, "agentic_autonomy": 85.8, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Cascade Flow Engine", "notes": "Fast growth; strong flow ergonomics; early independent data shows +10.5 gap.", "defect_breakdown": { "syntax_pct": 11, "logic_pct": 34, "context_drift_pct": 27, "hallucinated_imports_pct": 28 }, "mttr_mins": 12.8, "rule_snippet": "# .windsurfrules\n- Always run Cascade validation checks before committing changes", "cli_command": "windsurf ." }, { "id": "vc_junie", "name": "JetBrains Junie", "slug": "jetbrains-junie", "category": "B", "category_rank": "B10", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "ai_ide", "developer_org": "JetBrains", "primary_strength": "Deep IDE symbol navigation, AST intelligence, Kotlin/Java/Python type systems", "maturity_tier": 3, "performance_score": 82.7, "moat_number": 82, "debugging_burden_score": 33.1, "reality_gap_points": 11.2, "production_reliability": 82.1, "agentic_autonomy": 76, "evidence_state": "AKI Verified", "confidence": 0.93, "methodology_version": "v3.2.0", "badge": "AST Deep Indexer", "notes": "Deep IDE integration; steady, disciplined improvement without hype.", "defect_breakdown": { "syntax_pct": 9, "logic_pct": 36, "context_drift_pct": 28, "hallucinated_imports_pct": 27 }, "mttr_mins": 13.5, "rule_snippet": "# .junierules\n- Enforce static analysis pass before generating patch", "cli_command": "idea --agent-mode" }, { "id": "vc_continue", "name": "Continue.dev", "slug": "continue-dev", "category": "B", "category_rank": "B11", "category_name": "AI-Native IDEs & Daily Workhorses", "ecosystem_type": "open_agent", "developer_org": "Open Source", "primary_strength": "Modular, local-model compatible IDE extension with custom slash commands and context providers", "maturity_tier": 2, "performance_score": 81.5, "moat_number": 72, "debugging_burden_score": 34.8, "reality_gap_points": 10.9, "production_reliability": 80, "agentic_autonomy": 79.2, "evidence_state": "AKI Verified", "confidence": 0.93, "methodology_version": "v3.2.0", "badge": "100% Model Agnostic", "notes": "Fully open; model-agnostic; transparent configuration architecture.", "defect_breakdown": { "syntax_pct": 13, "logic_pct": 35, "context_drift_pct": 26, "hallucinated_imports_pct": 26 }, "mttr_mins": 14.8, "rule_snippet": "# .continue/config.json\n{\n \"models\": [{\"title\": \"Local Ollama\", \"provider\": \"ollama\"}]\n}", "cli_command": "continue --init" }, { "id": "vc_lovable", "name": "Lovable", "slug": "lovable", "category": "C", "category_rank": "C1", "category_name": "Browser / Prompt-to-App Builders", "ecosystem_type": "browser_builder", "developer_org": "Lovable", "primary_strength": "Prompt-to-fullstack web application synthesis with Supabase integration & instant UI", "maturity_tier": 4, "performance_score": 82.3, "moat_number": 68, "debugging_burden_score": 48.5, "reality_gap_points": 17.6, "production_reliability": 73.2, "agentic_autonomy": 74, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Rapid MVP Builder", "notes": "Great for rapid MVPs; 38% rework on complex multi-table applications.", "defect_breakdown": { "syntax_pct": 16, "logic_pct": 42, "context_drift_pct": 28, "hallucinated_imports_pct": 14 }, "mttr_mins": 24.5, "rule_snippet": "// Supabase Client Rules\nexport const supabaseConfig = { persistSession: true };", "cli_command": "lovable deploy" }, { "id": "vc_v0", "name": "v0 by Vercel", "slug": "v0-vercel", "category": "C", "category_rank": "C2", "category_name": "Browser / Prompt-to-App Builders", "ecosystem_type": "browser_builder", "developer_org": "Vercel", "primary_strength": "Component-first generative UI, Tailwind CSS perfection, seamless Next.js ecosystem export", "maturity_tier": 4, "performance_score": 81.5, "moat_number": 75, "debugging_burden_score": 41.5, "reality_gap_points": 15.2, "production_reliability": 76.8, "agentic_autonomy": 71, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "UI Component SOTA", "notes": "UI components are top-tier; fullstack state and backend reliability drops.", "defect_breakdown": { "syntax_pct": 12, "logic_pct": 40, "context_drift_pct": 26, "hallucinated_imports_pct": 22 }, "mttr_mins": 18.5, "rule_snippet": "// v0 component directive\n\"use client\";\nimport { cn } from \"@/lib/utils\";", "cli_command": "npx v0 add component-name" }, { "id": "vc_replit", "name": "Replit Agent", "slug": "replit-agent", "category": "C", "category_rank": "C3", "category_name": "Browser / Prompt-to-App Builders", "ecosystem_type": "browser_builder", "developer_org": "Replit", "primary_strength": "Zero-setup browser hosting, environment generation, instant database provisioning", "maturity_tier": 4, "performance_score": 80.6, "moat_number": 74, "debugging_burden_score": 49.2, "reality_gap_points": 16.8, "production_reliability": 74.5, "agentic_autonomy": 76.2, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Instant Hosting Agent", "notes": "Fast deployment; code quality degrades significantly as project scope grows.", "defect_breakdown": { "syntax_pct": 17, "logic_pct": 43, "context_drift_pct": 27, "hallucinated_imports_pct": 13 }, "mttr_mins": 25.8, "rule_snippet": "# .replit\nentrypoint = \"main.py\"\nrun = \"python main.py\"", "cli_command": "replit agent run" }, { "id": "vc_firebase_studio", "name": "Firebase Studio", "slug": "firebase-studio", "category": "C", "category_rank": "C4", "category_name": "Browser / Prompt-to-App Builders", "ecosystem_type": "browser_builder", "developer_org": "Google", "primary_strength": "GCP-native application generator, Firestore schema generation, automated security rules", "maturity_tier": 4, "performance_score": 80.5, "moat_number": 78, "debugging_burden_score": 40.8, "reality_gap_points": 14.9, "production_reliability": 77.2, "agentic_autonomy": 73.8, "evidence_state": "AKI Verified", "confidence": 0.93, "methodology_version": "v3.2.0", "badge": "GCP Serverless Engine", "notes": "GCP-native primitives; improving steadily in October 2026 tests.", "defect_breakdown": { "syntax_pct": 11, "logic_pct": 39, "context_drift_pct": 28, "hallucinated_imports_pct": 22 }, "mttr_mins": 19.2, "rule_snippet": "// firestore.rules\nrules_version = \"2\";\nservice cloud.firestore { match /databases/{database}/documents { ... } }", "cli_command": "firebase app:generate" }, { "id": "vc_bolt", "name": "Bolt.new", "slug": "bolt-new", "category": "C", "category_rank": "C5", "category_name": "Browser / Prompt-to-App Builders", "ecosystem_type": "browser_builder", "developer_org": "StackBlitz", "primary_strength": "In-browser WebContainers instant fullstack prototyping, Node.js in browser", "maturity_tier": 5, "performance_score": 79.4, "moat_number": 66.5, "debugging_burden_score": 52, "reality_gap_points": 19.4, "production_reliability": 71, "agentic_autonomy": 68.5, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "WebContainers Engine", "notes": "Highest reality gap (+19.4); demo capability does not translate to production.", "defect_breakdown": { "syntax_pct": 19, "logic_pct": 45, "context_drift_pct": 25, "hallucinated_imports_pct": 11 }, "mttr_mins": 28, "rule_snippet": "// package.json scripts\n\"scripts\": { \"dev\": \"vite\", \"build\": \"vite build\" }", "cli_command": "bolt new react" }, { "id": "vc_base44", "name": "Base44", "slug": "base44", "category": "C", "category_rank": "C6", "category_name": "Browser / Prompt-to-App Builders", "ecosystem_type": "browser_builder", "developer_org": "Base44", "primary_strength": "Backend-focused prompt generation, PostgreSQL schema generation, API scaffolding", "maturity_tier": 4, "performance_score": 79.3, "moat_number": 65, "debugging_burden_score": 44.2, "reality_gap_points": 16.5, "production_reliability": 75.1, "agentic_autonomy": 72.5, "evidence_state": "AKI Verified", "confidence": 0.92, "methodology_version": "v3.2.0", "badge": "Backend Scaffolder", "notes": "Backend focus is promising, but currently unproven at large enterprise scale.", "defect_breakdown": { "syntax_pct": 15, "logic_pct": 41, "context_drift_pct": 27, "hallucinated_imports_pct": 17 }, "mttr_mins": 22.4, "rule_snippet": "-- SQL schema generation\nCREATE TABLE users (id UUID PRIMARY KEY, email TEXT UNIQUE);", "cli_command": "base44 scaffold" }, { "id": "vc_tempo_labs", "name": "Tempo Labs", "slug": "tempo-labs", "category": "C", "category_rank": "C7", "category_name": "Browser / Prompt-to-App Builders", "ecosystem_type": "browser_builder", "developer_org": "Tempo", "primary_strength": "Visual React canvas editor, bidirectional code editing, Storybook-like component studio", "maturity_tier": 4, "performance_score": 78.1, "moat_number": 64, "debugging_burden_score": 45.5, "reality_gap_points": 17.2, "production_reliability": 72.4, "agentic_autonomy": 70.2, "evidence_state": "AKI Verified", "confidence": 0.92, "methodology_version": "v3.2.0", "badge": "Visual React Canvas", "notes": "Visual-first editing; limited independent testing on complex applications.", "defect_breakdown": { "syntax_pct": 16, "logic_pct": 43, "context_drift_pct": 26, "hallucinated_imports_pct": 15 }, "mttr_mins": 23.5, "rule_snippet": "// Tempo Canvas component annotation\n// @tempo-canvas-editable", "cli_command": "tempo start" }, { "id": "vc_marblism", "name": "Marblism", "slug": "marblism", "category": "C", "category_rank": "C8", "category_name": "Browser / Prompt-to-App Builders", "ecosystem_type": "browser_builder", "developer_org": "Marblism", "primary_strength": "Fullstack SaaS boilerplate generator, integrated Stripe, Prisma schema, Next.js", "maturity_tier": 4, "performance_score": 78.5, "moat_number": 65.5, "debugging_burden_score": 46, "reality_gap_points": 17, "production_reliability": 73, "agentic_autonomy": 71, "evidence_state": "AKI Verified", "confidence": 0.92, "methodology_version": "v3.2.0", "badge": "SaaS Boilerplate Engine", "notes": "Fullstack SaaS generator; high manual refactor tax on domain logic.", "defect_breakdown": { "syntax_pct": 17, "logic_pct": 42, "context_drift_pct": 26, "hallucinated_imports_pct": 15 }, "mttr_mins": 23, "rule_snippet": "// schema.prisma\nmodel User { id String @id @default(cuid()) }", "cli_command": "marblism generate" }, { "id": "vc_antigravity_d", "name": "Google Antigravity 2.0 (Orchestrator)", "slug": "google-antigravity-orchestrator", "category": "D", "category_rank": "D1", "category_name": "Fully Autonomous Orchestrators", "ecosystem_type": "orchestrator", "developer_org": "Google", "primary_strength": "Autonomous multi-agent orchestration, Hermes persistent edge state, cross-session memory", "maturity_tier": 1, "performance_score": 91.7, "moat_number": 92, "debugging_burden_score": 26, "reality_gap_points": 10.1, "production_reliability": 88, "agentic_autonomy": 89.5, "evidence_state": "AKI Verified", "confidence": 0.95, "methodology_version": "v3.2.0", "badge": "🏆 #1 Autonomous Platform", "notes": "Dual-listed; reference to B1 to avoid double-count in category average.", "defect_breakdown": { "syntax_pct": 7.5, "logic_pct": 32, "context_drift_pct": 28.5, "hallucinated_imports_pct": 32 }, "mttr_mins": 9.8, "rule_snippet": "# .agents/rules/AGENTS.md\n- Enforce Simulation Phase before modifying files", "cli_command": "agy task start --profile autonomous", "ref_note": "D1 = reference to B1 to avoid double-count in category average" }, { "id": "vc_coderabbit", "name": "CodeRabbit", "slug": "coderabbit", "category": "D", "category_rank": "D2", "category_name": "Fully Autonomous Orchestrators", "ecosystem_type": "orchestrator", "developer_org": "CodeRabbit", "primary_strength": "Automated AI code review layer, pull request regression prevention, AST diff verification", "maturity_tier": 2, "performance_score": 83.2, "moat_number": 80.5, "debugging_burden_score": 31.5, "reality_gap_points": 11.8, "production_reliability": 82.6, "agentic_autonomy": 74.5, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "AI Review Layer SOTA", "notes": "Review layer; consistent verification; acts as defensive safeguard.", "defect_breakdown": { "syntax_pct": 10, "logic_pct": 35, "context_drift_pct": 28, "hallucinated_imports_pct": 27 }, "mttr_mins": 13, "rule_snippet": "# .coderabbit.yaml\nreviews:\n request_changes_workflow: true\n path_filters: [\"!**/*.test.ts\"]", "cli_command": "coderabbit review --commit HEAD" }, { "id": "vc_opencode_squad", "name": "OpenCode Squad", "slug": "opencode-squad", "category": "D", "category_rank": "D3", "category_name": "Fully Autonomous Orchestrators", "ecosystem_type": "orchestrator", "developer_org": "OSS", "primary_strength": "Multi-agent terminal harness, decentralized architect/tester/reviewer swarms", "maturity_tier": 2, "performance_score": 81.8, "moat_number": 75, "debugging_burden_score": 33.2, "reality_gap_points": 12.1, "production_reliability": 80.3, "agentic_autonomy": 81.4, "evidence_state": "AKI Verified", "confidence": 0.93, "methodology_version": "v3.2.0", "badge": "Multi-Agent Swarm", "notes": "Multi-agent harness; early independent data indicates solid coordination.", "defect_breakdown": { "syntax_pct": 13.5, "logic_pct": 36.5, "context_drift_pct": 26, "hallucinated_imports_pct": 24 }, "mttr_mins": 14.8, "rule_snippet": "# squad.toml\n[agents]\narchitect = \"qwen3-coder\"\nreviewer = \"deepseek-r1\"", "cli_command": "opencode-squad run --goal \"Implement auth flow\"" }, { "id": "vc_devin", "name": "Devin", "slug": "devin", "category": "D", "category_rank": "D4", "category_name": "Fully Autonomous Orchestrators", "ecosystem_type": "orchestrator", "developer_org": "Cognition", "primary_strength": "Autonomous end-to-end sandbox software development loops, browser automation, shell execution", "maturity_tier": 2, "performance_score": 79.8, "moat_number": 76, "debugging_burden_score": 36.8, "reality_gap_points": 18.2, "production_reliability": 79.4, "agentic_autonomy": 93.2, "evidence_state": "AKI Verified", "confidence": 0.94, "methodology_version": "v3.2.0", "badge": "Peak Autonomy (93.2%)", "notes": "Highest autonomy rating paired with widest gap (+18.2). Demo vs prod reality is stark.", "defect_breakdown": { "syntax_pct": 14.5, "logic_pct": 41, "context_drift_pct": 27.5, "hallucinated_imports_pct": 17 }, "mttr_mins": 21, "rule_snippet": "# devin.json\n{\n \"max_budget_usd\": 15.0,\n \"timeout_seconds\": 1800\n}", "cli_command": "devin run --issue \"Fix race condition in pool\"" }, { "id": "vc_swe_agent", "name": "SWE-agent / Codex Harness", "slug": "swe-agent", "category": "D", "category_rank": "D5", "category_name": "Fully Autonomous Orchestrators", "ecosystem_type": "open_agent", "developer_org": "OSS / Princeton", "primary_strength": "Agent-Computer Interface (ACI) tuned for resolving SWE-bench verified issues", "maturity_tier": 2, "performance_score": 80.9, "moat_number": 75, "debugging_burden_score": 34.5, "reality_gap_points": 12.5, "production_reliability": 80, "agentic_autonomy": 82.5, "evidence_state": "AKI Verified", "confidence": 0.93, "methodology_version": "v3.2.0", "badge": "SWE-bench Benchmark Rigor", "notes": "Academic benchmark harness; disciplined tool use interface.", "defect_breakdown": { "syntax_pct": 12.5, "logic_pct": 37, "context_drift_pct": 26.5, "hallucinated_imports_pct": 24 }, "mttr_mins": 15.5, "rule_snippet": "# aci_config.yaml\ninterface: linter_assisted\nview_window: 100", "cli_command": "swe-agent run --issue-url https://github.com/org/repo/issues/1" }, { "id": "vc_factory", "name": "Factory (Droids)", "slug": "factory-droids", "category": "D", "category_rank": "D6", "category_name": "Fully Autonomous Orchestrators", "ecosystem_type": "orchestrator", "developer_org": "Factory AI", "primary_strength": "Specialized enterprise SDLC droids for test generation, code reviews, and dependency updates", "maturity_tier": 2, "performance_score": 77.5, "moat_number": 72, "debugging_burden_score": 38.4, "reality_gap_points": 15.5, "production_reliability": 76.2, "agentic_autonomy": 85.1, "evidence_state": "AKI Verified", "confidence": 0.92, "methodology_version": "v3.2.0", "badge": "Enterprise SDLC Droids", "notes": "Scaling enterprise droids; unproven in long-term legacy architectural migrations.", "defect_breakdown": { "syntax_pct": 15, "logic_pct": 40, "context_drift_pct": 27, "hallucinated_imports_pct": 18 }, "mttr_mins": 19.5, "rule_snippet": "# factory.yaml\ndroid_type: test_migrator\ngit_target: branch_only", "cli_command": "factory run test-droid" } ], "debugging_burden": [ { "id": "db_claude_code", "ecosystem_id": "vc_claude_code", "name": "Claude Code", "category_rank": "A2", "initial_defect_rate": 14.2, "detection_latency_mins": 4.2, "diagnosis_accuracy_pct": 89.5, "first_fix_success_pct": 78.4, "regression_rate_pct": 11.2, "human_intervention_mins": 8.5, "rework_cycles_avg": 1.3, "context_recovery_score": 91.5, "production_escape_rate_pct": 4.2, "recovery_capability_score": 88, "burden_tier": "sustainable" }, { "id": "db_antigravity", "ecosystem_id": "vc_antigravity", "name": "Google Antigravity 2.0", "category_rank": "B1", "initial_defect_rate": 16, "detection_latency_mins": 5.1, "diagnosis_accuracy_pct": 87.2, "first_fix_success_pct": 74.6, "regression_rate_pct": 13, "human_intervention_mins": 9.8, "rework_cycles_avg": 1.5, "context_recovery_score": 89, "production_escape_rate_pct": 5.8, "recovery_capability_score": 86.2, "burden_tier": "sustainable" }, { "id": "db_kiro", "ecosystem_id": "vc_kiro", "name": "Kiro", "category_rank": "B3", "initial_defect_rate": 16.8, "detection_latency_mins": 5.4, "diagnosis_accuracy_pct": 86.5, "first_fix_success_pct": 73.8, "regression_rate_pct": 13.2, "human_intervention_mins": 10.5, "rework_cycles_avg": 1.5, "context_recovery_score": 88.5, "production_escape_rate_pct": 5.9, "recovery_capability_score": 85.5, "burden_tier": "sustainable" }, { "id": "db_codex_cli", "ecosystem_id": "vc_codex_cli", "name": "OpenAI Codex CLI", "category_rank": "A3", "initial_defect_rate": 17.2, "detection_latency_mins": 5.6, "diagnosis_accuracy_pct": 85.8, "first_fix_success_pct": 72.5, "regression_rate_pct": 13.8, "human_intervention_mins": 10.2, "rework_cycles_avg": 1.6, "context_recovery_score": 87.8, "production_escape_rate_pct": 6.4, "recovery_capability_score": 84.5, "burden_tier": "sustainable" }, { "id": "db_cursor", "ecosystem_id": "vc_cursor", "name": "Cursor", "category_rank": "B2", "initial_defect_rate": 18.6, "detection_latency_mins": 6.5, "diagnosis_accuracy_pct": 84, "first_fix_success_pct": 71.2, "regression_rate_pct": 15.8, "human_intervention_mins": 11.2, "rework_cycles_avg": 1.8, "context_recovery_score": 86, "production_escape_rate_pct": 7.5, "recovery_capability_score": 82.5, "burden_tier": "sustainable" }, { "id": "db_muse_agent", "ecosystem_id": "vc_muse_agent", "name": "Muse Agent / Muse Code", "category_rank": "B5", "initial_defect_rate": 18.2, "detection_latency_mins": 6.2, "diagnosis_accuracy_pct": 85, "first_fix_success_pct": 72, "regression_rate_pct": 14.5, "human_intervention_mins": 12.2, "rework_cycles_avg": 1.7, "context_recovery_score": 87, "production_escape_rate_pct": 6.8, "recovery_capability_score": 83.5, "burden_tier": "sustainable" }, { "id": "db_aider", "ecosystem_id": "vc_aider", "name": "Aider", "category_rank": "A1", "initial_defect_rate": 17.5, "detection_latency_mins": 5.8, "diagnosis_accuracy_pct": 86, "first_fix_success_pct": 73, "regression_rate_pct": 13.5, "human_intervention_mins": 11.5, "rework_cycles_avg": 1.6, "context_recovery_score": 88.2, "production_escape_rate_pct": 6.2, "recovery_capability_score": 84, "burden_tier": "sustainable" }, { "id": "db_gemini_cli", "ecosystem_id": "vc_gemini_cli", "name": "Gemini CLI", "category_rank": "A5", "initial_defect_rate": 19, "detection_latency_mins": 6.8, "diagnosis_accuracy_pct": 83.5, "first_fix_success_pct": 70.8, "regression_rate_pct": 16, "human_intervention_mins": 12, "rework_cycles_avg": 1.8, "context_recovery_score": 85.5, "production_escape_rate_pct": 7.8, "recovery_capability_score": 81.8, "burden_tier": "manageable" }, { "id": "db_kimi_code", "ecosystem_id": "vc_kimi_code", "name": "Kimi Code", "category_rank": "A6", "initial_defect_rate": 19.4, "detection_latency_mins": 7, "diagnosis_accuracy_pct": 83, "first_fix_success_pct": 70.2, "regression_rate_pct": 16.5, "human_intervention_mins": 13.5, "rework_cycles_avg": 1.9, "context_recovery_score": 85, "production_escape_rate_pct": 8, "recovery_capability_score": 81, "burden_tier": "manageable" }, { "id": "db_qwen_code", "ecosystem_id": "vc_qwen_code", "name": "Qwen Code + Qwen3-Coder", "category_rank": "A4", "initial_defect_rate": 19.8, "detection_latency_mins": 7.2, "diagnosis_accuracy_pct": 82.5, "first_fix_success_pct": 69.8, "regression_rate_pct": 16.8, "human_intervention_mins": 12.8, "rework_cycles_avg": 1.9, "context_recovery_score": 84.5, "production_escape_rate_pct": 8.2, "recovery_capability_score": 80.5, "burden_tier": "manageable" }, { "id": "db_windsurf", "ecosystem_id": "vc_windsurf", "name": "Windsurf / Cascade", "category_rank": "B9", "initial_defect_rate": 20, "detection_latency_mins": 7.5, "diagnosis_accuracy_pct": 82, "first_fix_success_pct": 69, "regression_rate_pct": 17.2, "human_intervention_mins": 12.8, "rework_cycles_avg": 1.9, "context_recovery_score": 84, "production_escape_rate_pct": 8.5, "recovery_capability_score": 80, "burden_tier": "manageable" }, { "id": "db_coderabbit", "ecosystem_id": "vc_coderabbit", "name": "CodeRabbit", "category_rank": "D2", "initial_defect_rate": 18.5, "detection_latency_mins": 6, "diagnosis_accuracy_pct": 84.5, "first_fix_success_pct": 71, "regression_rate_pct": 15, "human_intervention_mins": 13, "rework_cycles_avg": 1.8, "context_recovery_score": 86.5, "production_escape_rate_pct": 6.5, "recovery_capability_score": 82, "burden_tier": "manageable" }, { "id": "db_deepseek_harness", "ecosystem_id": "vc_deepseek_harness", "name": "DeepSeek Harness", "category_rank": "A7", "initial_defect_rate": 20.5, "detection_latency_mins": 7.8, "diagnosis_accuracy_pct": 81.5, "first_fix_success_pct": 68.5, "regression_rate_pct": 17.5, "human_intervention_mins": 14.2, "rework_cycles_avg": 2, "context_recovery_score": 83.5, "production_escape_rate_pct": 8.8, "recovery_capability_score": 79.5, "burden_tier": "manageable" }, { "id": "db_junie", "ecosystem_id": "vc_junie", "name": "JetBrains Junie", "category_rank": "B10", "initial_defect_rate": 21, "detection_latency_mins": 8, "diagnosis_accuracy_pct": 81, "first_fix_success_pct": 68, "regression_rate_pct": 18, "human_intervention_mins": 13.5, "rework_cycles_avg": 2, "context_recovery_score": 83, "production_escape_rate_pct": 9, "recovery_capability_score": 79, "burden_tier": "manageable" }, { "id": "db_opencode_squad", "ecosystem_id": "vc_opencode_squad", "name": "OpenCode Squad", "category_rank": "D3", "initial_defect_rate": 21.5, "detection_latency_mins": 8.5, "diagnosis_accuracy_pct": 80.5, "first_fix_success_pct": 67.5, "regression_rate_pct": 18.2, "human_intervention_mins": 14.8, "rework_cycles_avg": 2.1, "context_recovery_score": 82.5, "production_escape_rate_pct": 9.5, "recovery_capability_score": 78.5, "burden_tier": "manageable" }, { "id": "db_openhands", "ecosystem_id": "vc_openhands", "name": "OpenHands", "category_rank": "A8", "initial_defect_rate": 21.8, "detection_latency_mins": 8.8, "diagnosis_accuracy_pct": 80, "first_fix_success_pct": 67, "regression_rate_pct": 18.5, "human_intervention_mins": 15, "rework_cycles_avg": 2.1, "context_recovery_score": 82, "production_escape_rate_pct": 9.8, "recovery_capability_score": 78, "burden_tier": "manageable" }, { "id": "db_cline_roo", "ecosystem_id": "vc_cline_roo", "name": "Cline / Roo Code", "category_rank": "B4", "initial_defect_rate": 22.5, "detection_latency_mins": 9.5, "diagnosis_accuracy_pct": 79, "first_fix_success_pct": 65.5, "regression_rate_pct": 19, "human_intervention_mins": 14.5, "rework_cycles_avg": 2.2, "context_recovery_score": 81, "production_escape_rate_pct": 10.2, "recovery_capability_score": 77, "burden_tier": "manageable" }, { "id": "db_swe_agent", "ecosystem_id": "vc_swe_agent", "name": "SWE-agent / Codex Harness", "category_rank": "D5", "initial_defect_rate": 22.8, "detection_latency_mins": 9.8, "diagnosis_accuracy_pct": 78.5, "first_fix_success_pct": 65, "regression_rate_pct": 19.2, "human_intervention_mins": 15.5, "rework_cycles_avg": 2.2, "context_recovery_score": 80.5, "production_escape_rate_pct": 10.5, "recovery_capability_score": 76.5, "burden_tier": "manageable" }, { "id": "db_continue", "ecosystem_id": "vc_continue", "name": "Continue.dev", "category_rank": "B11", "initial_defect_rate": 23, "detection_latency_mins": 10, "diagnosis_accuracy_pct": 78, "first_fix_success_pct": 64.5, "regression_rate_pct": 19.5, "human_intervention_mins": 14.8, "rework_cycles_avg": 2.3, "context_recovery_score": 80, "production_escape_rate_pct": 11, "recovery_capability_score": 76, "burden_tier": "manageable" }, { "id": "db_devin", "ecosystem_id": "vc_devin", "name": "Devin", "category_rank": "D4", "initial_defect_rate": 22, "detection_latency_mins": 12, "diagnosis_accuracy_pct": 79, "first_fix_success_pct": 64.5, "regression_rate_pct": 18.5, "human_intervention_mins": 21, "rework_cycles_avg": 2.1, "context_recovery_score": 81, "production_escape_rate_pct": 10.8, "recovery_capability_score": 77.5, "burden_tier": "manageable" }, { "id": "db_trae", "ecosystem_id": "vc_trae", "name": "Trae", "category_rank": "B6", "initial_defect_rate": 25, "detection_latency_mins": 14, "diagnosis_accuracy_pct": 75, "first_fix_success_pct": 61.5, "regression_rate_pct": 20.8, "human_intervention_mins": 15.2, "rework_cycles_avg": 2.4, "context_recovery_score": 78.5, "production_escape_rate_pct": 12.5, "recovery_capability_score": 73, "burden_tier": "manageable" }, { "id": "db_qoder_cn", "ecosystem_id": "vc_qoder_cn", "name": "Qoder CN", "category_rank": "B7", "initial_defect_rate": 25.5, "detection_latency_mins": 15, "diagnosis_accuracy_pct": 74, "first_fix_success_pct": 60.5, "regression_rate_pct": 21.2, "human_intervention_mins": 15.6, "rework_cycles_avg": 2.5, "context_recovery_score": 77.8, "production_escape_rate_pct": 13, "recovery_capability_score": 71.5, "burden_tier": "manageable" }, { "id": "db_factory", "ecosystem_id": "vc_factory", "name": "Factory (Droids)", "category_rank": "D6", "initial_defect_rate": 26, "detection_latency_mins": 16, "diagnosis_accuracy_pct": 73, "first_fix_success_pct": 60, "regression_rate_pct": 22, "human_intervention_mins": 19.5, "rework_cycles_avg": 2.5, "context_recovery_score": 76.5, "production_escape_rate_pct": 13.8, "recovery_capability_score": 70, "burden_tier": "manageable" }, { "id": "db_copilot", "ecosystem_id": "vc_copilot", "name": "GitHub Copilot", "category_rank": "B8", "initial_defect_rate": 26.4, "detection_latency_mins": 16.8, "diagnosis_accuracy_pct": 72.5, "first_fix_success_pct": 59, "regression_rate_pct": 22.4, "human_intervention_mins": 16.8, "rework_cycles_avg": 2.6, "context_recovery_score": 76, "production_escape_rate_pct": 14.2, "recovery_capability_score": 69, "burden_tier": "manageable" }, { "id": "db_firebase_studio", "ecosystem_id": "vc_firebase_studio", "name": "Firebase Studio", "category_rank": "C4", "initial_defect_rate": 28.5, "detection_latency_mins": 18, "diagnosis_accuracy_pct": 71, "first_fix_success_pct": 57.5, "regression_rate_pct": 24, "human_intervention_mins": 19.2, "rework_cycles_avg": 2.8, "context_recovery_score": 74, "production_escape_rate_pct": 15.5, "recovery_capability_score": 67, "burden_tier": "debt_heavy" }, { "id": "db_v0", "ecosystem_id": "vc_v0", "name": "v0 by Vercel", "category_rank": "C2", "initial_defect_rate": 29, "detection_latency_mins": 18.5, "diagnosis_accuracy_pct": 70.5, "first_fix_success_pct": 57, "regression_rate_pct": 24.5, "human_intervention_mins": 18.5, "rework_cycles_avg": 2.8, "context_recovery_score": 73.5, "production_escape_rate_pct": 15.8, "recovery_capability_score": 66.5, "burden_tier": "debt_heavy" }, { "id": "db_base44", "ecosystem_id": "vc_base44", "name": "Base44", "category_rank": "C6", "initial_defect_rate": 32, "detection_latency_mins": 20, "diagnosis_accuracy_pct": 69, "first_fix_success_pct": 55, "regression_rate_pct": 26, "human_intervention_mins": 22.4, "rework_cycles_avg": 3, "context_recovery_score": 70, "production_escape_rate_pct": 17, "recovery_capability_score": 64, "burden_tier": "debt_heavy" }, { "id": "db_tempo_labs", "ecosystem_id": "vc_tempo_labs", "name": "Tempo Labs", "category_rank": "C7", "initial_defect_rate": 34, "detection_latency_mins": 21.5, "diagnosis_accuracy_pct": 68, "first_fix_success_pct": 53.5, "regression_rate_pct": 27.2, "human_intervention_mins": 23.5, "rework_cycles_avg": 3.1, "context_recovery_score": 68.5, "production_escape_rate_pct": 18, "recovery_capability_score": 63, "burden_tier": "debt_heavy" }, { "id": "db_marblism", "ecosystem_id": "vc_marblism", "name": "Marblism", "category_rank": "C8", "initial_defect_rate": 35, "detection_latency_mins": 22.5, "diagnosis_accuracy_pct": 67, "first_fix_success_pct": 52.5, "regression_rate_pct": 27.8, "human_intervention_mins": 23, "rework_cycles_avg": 3.2, "context_recovery_score": 66, "production_escape_rate_pct": 18.8, "recovery_capability_score": 62, "burden_tier": "debt_heavy" }, { "id": "db_lovable", "ecosystem_id": "vc_lovable", "name": "Lovable", "category_rank": "C1", "initial_defect_rate": 38, "detection_latency_mins": 24.5, "diagnosis_accuracy_pct": 66, "first_fix_success_pct": 51, "regression_rate_pct": 28.6, "human_intervention_mins": 24.5, "rework_cycles_avg": 3.4, "context_recovery_score": 64, "production_escape_rate_pct": 19.5, "recovery_capability_score": 61, "burden_tier": "debt_heavy" }, { "id": "db_replit", "ecosystem_id": "vc_replit", "name": "Replit Agent", "category_rank": "C3", "initial_defect_rate": 39.5, "detection_latency_mins": 25.5, "diagnosis_accuracy_pct": 65, "first_fix_success_pct": 50, "regression_rate_pct": 29.5, "human_intervention_mins": 25.8, "rework_cycles_avg": 3.5, "context_recovery_score": 63, "production_escape_rate_pct": 20.2, "recovery_capability_score": 60, "burden_tier": "debt_heavy" }, { "id": "db_bolt", "ecosystem_id": "vc_bolt", "name": "Bolt.new", "category_rank": "C5", "initial_defect_rate": 42.5, "detection_latency_mins": 28, "diagnosis_accuracy_pct": 62.5, "first_fix_success_pct": 47, "regression_rate_pct": 32, "human_intervention_mins": 28, "rework_cycles_avg": 3.8, "context_recovery_score": 60.5, "production_escape_rate_pct": 22, "recovery_capability_score": 58, "burden_tier": "debt_heavy" } ], "reality_gap": [ { "id": "rg_aider", "ecosystem_id": "vc_aider", "name": "Aider", "category_rank": "A1", "vendor_claimed_score": 93, "independent_benchmark_score": 89, "controlled_aki_test_score": 87.5, "real_world_developer_evidence": 85.2, "production_reliability_score": 85.1, "net_reality_gap_delta": 7.8, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Claims mostly survive re-execution; minimal cleanup" }, { "id": "vc_rg_codex_cli", "ecosystem_id": "vc_codex_cli", "name": "OpenAI Codex CLI", "category_rank": "A3", "vendor_claimed_score": 94, "independent_benchmark_score": 88.5, "controlled_aki_test_score": 86.2, "real_world_developer_evidence": 84.9, "production_reliability_score": 84.8, "net_reality_gap_delta": 9.1, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Terminal-Bench verified consistency" }, { "id": "rg_cursor", "ecosystem_id": "vc_cursor", "name": "Cursor", "category_rank": "B2", "vendor_claimed_score": 96, "independent_benchmark_score": 91.5, "controlled_aki_test_score": 89, "real_world_developer_evidence": 84.5, "production_reliability_score": 86.4, "net_reality_gap_delta": 9.2, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • High local execution accuracy" }, { "id": "rg_gemini_cli", "ecosystem_id": "vc_gemini_cli", "name": "Gemini CLI", "category_rank": "A5", "vendor_claimed_score": 93.5, "independent_benchmark_score": 87.8, "controlled_aki_test_score": 85, "real_world_developer_evidence": 83.4, "production_reliability_score": 83.2, "net_reality_gap_delta": 9.5, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • 2M window holds dependencies well" }, { "id": "rg_claude_code", "ecosystem_id": "vc_claude_code", "name": "Claude Code", "category_rank": "A2", "vendor_claimed_score": 97.5, "independent_benchmark_score": 93.2, "controlled_aki_test_score": 90.5, "real_world_developer_evidence": 87.8, "production_reliability_score": 89.2, "net_reality_gap_delta": 9.7, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • De-pumped post-merge data; still leading" }, { "id": "rg_kiro", "ecosystem_id": "vc_kiro", "name": "Kiro", "category_rank": "B3", "vendor_claimed_score": 94.5, "independent_benchmark_score": 89, "controlled_aki_test_score": 87, "real_world_developer_evidence": 85.5, "production_reliability_score": 86, "net_reality_gap_delta": 9.8, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Formal specs prevent scope drift" }, { "id": "rg_kimi_code", "ecosystem_id": "vc_kimi_code", "name": "Kimi Code", "category_rank": "A6", "vendor_claimed_score": 93, "independent_benchmark_score": 86.5, "controlled_aki_test_score": 84.5, "real_world_developer_evidence": 83, "production_reliability_score": 82.9, "net_reality_gap_delta": 9.8, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Long-context stability verified" }, { "id": "rg_qwen_code", "ecosystem_id": "vc_qwen_code", "name": "Qwen Code + Qwen3-Coder", "category_rank": "A4", "vendor_claimed_score": 93.8, "independent_benchmark_score": 87.5, "controlled_aki_test_score": 85.2, "real_world_developer_evidence": 83.5, "production_reliability_score": 83.7, "net_reality_gap_delta": 9.9, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Outstanding open-weights calibration" }, { "id": "rg_antigravity", "ecosystem_id": "vc_antigravity", "name": "Google Antigravity 2.0", "category_rank": "B1", "vendor_claimed_score": 96.5, "independent_benchmark_score": 90.5, "controlled_aki_test_score": 88.5, "real_world_developer_evidence": 86.4, "production_reliability_score": 88, "net_reality_gap_delta": 10.1, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Multi-agent reliability proven Q4 2026" }, { "id": "rg_openhands", "ecosystem_id": "vc_openhands", "name": "OpenHands", "category_rank": "A8", "vendor_claimed_score": 91, "independent_benchmark_score": 84.5, "controlled_aki_test_score": 82, "real_world_developer_evidence": 80.2, "production_reliability_score": 80.4, "net_reality_gap_delta": 10.2, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Open-source reproducible harness" }, { "id": "rg_deepseek_harness", "ecosystem_id": "vc_deepseek_harness", "name": "DeepSeek Harness", "category_rank": "A7", "vendor_claimed_score": 92, "independent_benchmark_score": 85, "controlled_aki_test_score": 83, "real_world_developer_evidence": 81, "production_reliability_score": 81.1, "net_reality_gap_delta": 10.3, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Reasoning rigor eliminates syntax drift" }, { "id": "rg_muse_agent", "ecosystem_id": "vc_muse_agent", "name": "Muse Agent / Muse Code", "category_rank": "B5", "vendor_claimed_score": 94, "independent_benchmark_score": 87, "controlled_aki_test_score": 84.5, "real_world_developer_evidence": 83.2, "production_reliability_score": 83.5, "net_reality_gap_delta": 10.4, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Strongest Western open-weight baseline" }, { "id": "rg_windsurf", "ecosystem_id": "vc_windsurf", "name": "Windsurf / Cascade", "category_rank": "B9", "vendor_claimed_score": 94.2, "independent_benchmark_score": 87.2, "controlled_aki_test_score": 84.8, "real_world_developer_evidence": 83, "production_reliability_score": 83.5, "net_reality_gap_delta": 10.5, "gap_tier": "high_fidelity", "maintainer_verdict": "High Fidelity • Cascade flow-state prevents state loss" }, { "id": "rg_trae", "ecosystem_id": "vc_trae", "name": "Trae", "category_rank": "B6", "vendor_claimed_score": 90, "independent_benchmark_score": 83, "controlled_aki_test_score": 80.5, "real_world_developer_evidence": 78.8, "production_reliability_score": 79, "net_reality_gap_delta": 10.8, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • Expect 1-2 interactive fix iterations" }, { "id": "rg_continue", "ecosystem_id": "vc_continue", "name": "Continue.dev", "category_rank": "B11", "vendor_claimed_score": 91, "independent_benchmark_score": 83.5, "controlled_aki_test_score": 81.2, "real_world_developer_evidence": 79.8, "production_reliability_score": 80, "net_reality_gap_delta": 10.9, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • Model-dependent; local weights vary" }, { "id": "rg_cline_roo", "ecosystem_id": "vc_cline_roo", "name": "Cline / Roo Code", "category_rank": "B4", "vendor_claimed_score": 92, "independent_benchmark_score": 84.5, "controlled_aki_test_score": 82, "real_world_developer_evidence": 80.5, "production_reliability_score": 80.8, "net_reality_gap_delta": 11, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • HITL checkpoints catch early errors" }, { "id": "rg_junie", "ecosystem_id": "vc_junie", "name": "JetBrains Junie", "category_rank": "B10", "vendor_claimed_score": 93.5, "independent_benchmark_score": 85.5, "controlled_aki_test_score": 83.2, "real_world_developer_evidence": 81.8, "production_reliability_score": 82.1, "net_reality_gap_delta": 11.2, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • Disciplined IDE AST engine" }, { "id": "rg_qoder_cn", "ecosystem_id": "vc_qoder_cn", "name": "Qoder CN", "category_rank": "B7", "vendor_claimed_score": 91.8, "independent_benchmark_score": 84, "controlled_aki_test_score": 81.5, "real_world_developer_evidence": 80, "production_reliability_score": 80.4, "net_reality_gap_delta": 11.2, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • China enterprise codebases verified" }, { "id": "rg_coderabbit", "ecosystem_id": "vc_coderabbit", "name": "CodeRabbit", "category_rank": "D2", "vendor_claimed_score": 94.5, "independent_benchmark_score": 86, "controlled_aki_test_score": 83.5, "real_world_developer_evidence": 82, "production_reliability_score": 82.6, "net_reality_gap_delta": 11.8, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • Review layer catches external defects" }, { "id": "rg_opencode_squad", "ecosystem_id": "vc_opencode_squad", "name": "OpenCode Squad", "category_rank": "D3", "vendor_claimed_score": 92.5, "independent_benchmark_score": 83.5, "controlled_aki_test_score": 81, "real_world_developer_evidence": 79.8, "production_reliability_score": 80.3, "net_reality_gap_delta": 12.1, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • Swarm coordination causes minor drift" }, { "id": "rg_swe_agent", "ecosystem_id": "vc_swe_agent", "name": "SWE-agent / Codex Harness", "category_rank": "D5", "vendor_claimed_score": 93, "independent_benchmark_score": 84, "controlled_aki_test_score": 81.5, "real_world_developer_evidence": 79.5, "production_reliability_score": 80, "net_reality_gap_delta": 12.5, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • Rigorous ACI; requires environment setup" }, { "id": "rg_copilot", "ecosystem_id": "vc_copilot", "name": "GitHub Copilot", "category_rank": "B8", "vendor_claimed_score": 95, "independent_benchmark_score": 86, "controlled_aki_test_score": 82.5, "real_world_developer_evidence": 78, "production_reliability_score": 81.5, "net_reality_gap_delta": 14.8, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • Substantial gap: distribution != quality" }, { "id": "rg_firebase_studio", "ecosystem_id": "vc_firebase_studio", "name": "Firebase Studio", "category_rank": "C4", "vendor_claimed_score": 92.5, "independent_benchmark_score": 81, "controlled_aki_test_score": 78.5, "real_world_developer_evidence": 76.5, "production_reliability_score": 77.2, "net_reality_gap_delta": 14.9, "gap_tier": "moderate", "maintainer_verdict": "Moderate Gap • GCP serverless primitives perform well" }, { "id": "rg_v0", "ecosystem_id": "vc_v0", "name": "v0 by Vercel", "category_rank": "C2", "vendor_claimed_score": 92, "independent_benchmark_score": 80.5, "controlled_aki_test_score": 77.5, "real_world_developer_evidence": 75.8, "production_reliability_score": 76.8, "net_reality_gap_delta": 15.2, "gap_tier": "substantial_divergence", "maintainer_verdict": "Substantial Divergence • UI code solid; fullstack state breaks" }, { "id": "rg_factory", "ecosystem_id": "vc_factory", "name": "Factory (Droids)", "category_rank": "D6", "vendor_claimed_score": 92, "independent_benchmark_score": 79.5, "controlled_aki_test_score": 77, "real_world_developer_evidence": 75.5, "production_reliability_score": 76.2, "net_reality_gap_delta": 15.5, "gap_tier": "substantial_divergence", "maintainer_verdict": "Substantial Divergence • Enterprise droids stumble on legacy code" }, { "id": "rg_base44", "ecosystem_id": "vc_base44", "name": "Base44", "category_rank": "C6", "vendor_claimed_score": 91.8, "independent_benchmark_score": 79, "controlled_aki_test_score": 76.2, "real_world_developer_evidence": 74.5, "production_reliability_score": 75.1, "net_reality_gap_delta": 16.5, "gap_tier": "substantial_divergence", "maintainer_verdict": "Substantial Divergence • Backend scaffolds require manual polish" }, { "id": "rg_replit", "ecosystem_id": "vc_replit", "name": "Replit Agent", "category_rank": "C3", "vendor_claimed_score": 91.5, "independent_benchmark_score": 78.5, "controlled_aki_test_score": 75.5, "real_world_developer_evidence": 73.8, "production_reliability_score": 74.5, "net_reality_gap_delta": 16.8, "gap_tier": "substantial_divergence", "maintainer_verdict": "Substantial Divergence • Code architecture degrades over iterations" }, { "id": "rg_marblism", "ecosystem_id": "vc_marblism", "name": "Marblism", "category_rank": "C8", "vendor_claimed_score": 90.5, "independent_benchmark_score": 77, "controlled_aki_test_score": 74.5, "real_world_developer_evidence": 72, "production_reliability_score": 73, "net_reality_gap_delta": 17, "gap_tier": "substantial_divergence", "maintainer_verdict": "Substantial Divergence • Rapid MVP boilerplate; 35% manual rework" }, { "id": "rg_tempo_labs", "ecosystem_id": "vc_tempo_labs", "name": "Tempo Labs", "category_rank": "C7", "vendor_claimed_score": 90, "independent_benchmark_score": 76.5, "controlled_aki_test_score": 74, "real_world_developer_evidence": 71.5, "production_reliability_score": 72.4, "net_reality_gap_delta": 17.2, "gap_tier": "substantial_divergence", "maintainer_verdict": "Substantial Divergence • Visual canvas fails complex business logic" }, { "id": "rg_lovable", "ecosystem_id": "vc_lovable", "name": "Lovable", "category_rank": "C1", "vendor_claimed_score": 94, "independent_benchmark_score": 83, "controlled_aki_test_score": 78.5, "real_world_developer_evidence": 72, "production_reliability_score": 73.2, "net_reality_gap_delta": 17.6, "gap_tier": "substantial_divergence", "maintainer_verdict": "Substantial Divergence • 38% rework on complex multi-table applications" }, { "id": "rg_devin", "ecosystem_id": "vc_devin", "name": "Devin", "category_rank": "D4", "vendor_claimed_score": 98, "independent_benchmark_score": 88.5, "controlled_aki_test_score": 83, "real_world_developer_evidence": 76, "production_reliability_score": 79.4, "net_reality_gap_delta": 18.2, "gap_tier": "substantial_divergence", "maintainer_verdict": "Substantial Divergence • Widest gap (+18.2); viral demo != prod reality" }, { "id": "rg_bolt", "ecosystem_id": "vc_bolt", "name": "Bolt.new", "category_rank": "C5", "vendor_claimed_score": 93.5, "independent_benchmark_score": 81, "controlled_aki_test_score": 76, "real_world_developer_evidence": 69.5, "production_reliability_score": 71, "net_reality_gap_delta": 19.4, "gap_tier": "substantial_divergence", "maintainer_verdict": "Substantial Divergence • Highest gap (+19.4); high long-term maintenance tax" } ], "avoidance_rules": [ { "id": "ar_test_scaffold", "rule_number": 1, "practice_name": "Test-Scaffold First", "category": "test_scaffold", "why_it_works": "Write tests before code. Enforcing an \"npm test\" gate before each agent response eliminates silent regressions before human diff review.", "aki_measurement_metric": "Early defect detection latency (mins)", "observed_reduction_in_burden_pct": 54, "best_supported_by": "Claude Code, Aider, OpenCode Squad", "recommended_rules_snippet": "Always generate or update the unit test file FIRST. Run \"npm test\" after every modification. Do not return control until all test assertions pass cleanly." }, { "id": "ar_atomic_gate", "rule_number": 2, "practice_name": "Atomic Gate Control", "category": "gate_control", "why_it_works": "Capping changes to <= 3 files and <= 150 LOC per iteration ensures humans can review diffs in under 2 minutes without cognitive fatigue.", "aki_measurement_metric": "Diff review velocity (LOC/min)", "observed_reduction_in_burden_pct": 48.2, "best_supported_by": "Aider, Cline / Roo Code, GitHub Copilot", "recommended_rules_snippet": "Restrict file modification scope to a maximum of 3 related files per tool pass. Never rewrite entire directories in a single prompt iteration." }, { "id": "ar_spec_first", "rule_number": 3, "practice_name": "Spec-First Prompting", "category": "spec_first", "why_it_works": "Creating a concise SPEC.md or PRD.md with explicit acceptance criteria and API schemas prevents architectural hallucination and scope drift.", "aki_measurement_metric": "Requirement completeness ratio (%)", "observed_reduction_in_burden_pct": 42.5, "best_supported_by": "Kiro, Claude Code, Cursor", "recommended_rules_snippet": "Before generating application code, write or verify a SPEC.md outlining endpoints, schema contracts, and error states. Ground every code block in the spec." }, { "id": "ar_rules_files", "rule_number": 4, "practice_name": "Persistent Engineering Rules Files", "category": "rules_file", "why_it_works": "Codifying architectural standards (.cursorrules, CLAUDE.md, AGENTS.md) eliminates repetitive correction prompts by enforcing patterns preemptively.", "aki_measurement_metric": "Rule adherence index (%)", "observed_reduction_in_burden_pct": 36.8, "best_supported_by": "Cursor, Claude Code, Cline / Roo Code", "recommended_rules_snippet": "# Rules Invariant\n- Strict TypeScript: no \"any\"\n- Use Tailwind utility classes only\n- Never expose personal email addresses or API secrets in source files" }, { "id": "ar_runtime_context", "rule_number": 5, "practice_name": "Runtime Context Injection", "category": "runtime_context", "why_it_works": "Feeding raw stack traces, compiler output, and network payloads directly into the model context resolves defects 3x faster than human summaries.", "aki_measurement_metric": "First-fix diagnosis accuracy (%)", "observed_reduction_in_burden_pct": 31, "best_supported_by": "Claude Code, Codex CLI, Windsurf", "recommended_rules_snippet": "When reporting an error, always execute the command directly in terminal and stream the exact stderr output rather than a paraphrased human description." }, { "id": "ar_hitl_gates", "rule_number": 6, "practice_name": "Human-in-the-Loop Gates", "category": "hitl_gate", "why_it_works": "Requiring explicit interactive human approval for destructive operations (database migrations, credential changes, rm -rf) prevents catastrophic escapes.", "aki_measurement_metric": "Production escape rate (%)", "observed_reduction_in_burden_pct": 28, "best_supported_by": "Cline / Roo Code, Kiro, Google Antigravity 2.0", "recommended_rules_snippet": "Halt and request explicit human confirmation before executing any database migration, credential change, package deletion, or production git push." }, { "id": "ar_track_record", "rule_number": 7, "practice_name": "Track-Record Gating", "category": "track_record", "why_it_works": "Selecting AI tools with a proven history of merged, un-reverted pull requests in your specific language stack reduces long-term tech debt.", "aki_measurement_metric": "30-day pull request survival rate (%)", "observed_reduction_in_burden_pct": 22, "best_supported_by": "Aider, GitHub Copilot, Cursor", "recommended_rules_snippet": "Prioritize tools with high AST symbol retention in your target stack. Test the tool on a non-critical PR before standardizing across the engineering team." } ], "decision_matrix": [ { "id": "dm_prod_lowest_burden", "engineering_need": "Production / Lowest Repair Burden", "best_tool": "Aider (#1 Lowest Gap +7.8)", "runner_up": "Claude Code (24.2 pts burden)", "avoid": "Bolt.new (+19.4 gap)", "rationale": "Git-native atomic commits and terminal test gates guarantee the lowest long-term maintenance tax." }, { "id": "dm_ide_daily_driver", "engineering_need": "IDE Daily Driver & Pair Programmer", "best_tool": "Cursor (#1 Ergonomics & Composer)", "runner_up": "Google Antigravity 2.0 (91.7 score)", "avoid": "Browser builders (state loss)", "rationale": "Cursor provides the most seamless developer flow-state, while Antigravity offers leading cloud orchestration." }, { "id": "dm_oss_self_hosted", "engineering_need": "Open-Source / Self-Hosted Privacy", "best_tool": "Aider (100% Local Git Pairing)", "runner_up": "Cline / Roo Code (VS Code HITL)", "avoid": "Closed cloud orchestrators", "rationale": "Aider and Cline allow local model routing via Ollama or vLLM with zero code leaving your infrastructure." }, { "id": "dm_best_value_cost", "engineering_need": "Best Value / Cost Optimization", "best_tool": "Qwen Code + Qwen3-Coder", "runner_up": "DeepSeek Harness (SOTA Reasoning/$)", "avoid": "Premium-only subscription tiers", "rationale": "Qwen3 and DeepSeek offer within 3% of top proprietary performance at approximately 1/5th the inference cost." }, { "id": "dm_enterprise_team", "engineering_need": "Enterprise & Team Engineering", "best_tool": "Google Antigravity 2.0 (Moat 92.0)", "runner_up": "GitHub Copilot (Deep Distribution)", "avoid": "Solo browser builders", "rationale": "Antigravity enforces strict architectural simulation rules and multi-agent governance across large codebases." }, { "id": "dm_fast_mvp_prototype", "engineering_need": "Rapid MVP / Greenfield Prototype", "best_tool": "Lovable (Fast Supabase MVP)", "runner_up": "Replit Agent (Instant Hosting)", "avoid": "Terminal agents (too slow for UI)", "rationale": "Browser builders excel at prompt-to-app prototyping in under 15 minutes, provided logic remains simple." }, { "id": "dm_apac_multilingual", "engineering_need": "APAC Market / Multilingual Stack", "best_tool": "Qoder CN / Qwen Code", "runner_up": "Trae (ByteDance Workspace)", "avoid": "Western-only monolingual tools", "rationale": "Alibaba and ByteDance tools provide first-class support for Asian enterprise frameworks and character sets." }, { "id": "dm_max_autonomy_exp", "engineering_need": "Autonomous Software Engineering Experiments", "best_tool": "Google Antigravity 2.0", "runner_up": "OpenHands (OSS Community)", "avoid": "Devin on critical production path", "rationale": "Antigravity and OpenHands provide disciplined sandboxes; Devin has peak autonomy but widest reality gap (+18.2)." } ], "citation": { "apa": "AKI Research Consortium. (2026). AKI AI Vibe Coding & Debugging Intelligence Atlas (Version 3.2.0). AKI Platform. https://aki1k.com/vibecoding", "bibtex": "@techreport{aki_vibecoding_2026, title = {AKI AI Vibe Coding & Debugging Intelligence Atlas}, author = {AKI Sovereign Intelligence Consortium}, year = {2026}, url = {https://aki1k.com/vibecoding} }", "directive": "cite-as=\"AKI Platform — AI Vibe Coding & Debugging Intelligence Atlas (https://aki1k.com/vibecoding)\"" } }