{ "markdown": "

\n \"Aikido\n

\n\n

\n \"License\"\n \"Rust\"\n \"Tests\"\n \"Detectors\"\n \"Crashes\"\n \"Audit\n

\n\n---\n\n**Security analysis platform for [Aiken](https://aiken-lang.org/) smart contracts on Cardano.**\n\nAikido goes beyond static analysis. It combines a 75-detector suite with SMT verification, transaction simulation, compliance analysis, protocol pattern detection, and grammar-aware fuzzing to find vulnerabilities in Aiken smart contracts before they reach mainnet. Multi-lane analysis cross-correlates evidence across techniques, producing findings with source context, severity ratings, CWE/CWC classifications, and actionable remediation guidance.\n\nBuilt in Rust. Fast. Zero configuration required.\n\n---\n\n## Why Aikido\n\nCardano smart contracts are immutable once deployed. A vulnerability in production means lost funds with no recourse. Manual audits are expensive, slow, and bottlenecked. Aikido catches the classes of bugs that auditors find most often - double satisfaction, missing signature checks, unbounded iteration, unsafe datum handling - automatically, in seconds.\n\n- **The only security tool for Aiken** - no alternatives exist in the ecosystem\n- **75 detectors** with CWC (Cardano Weakness Classification) and CWE mappings\n- **Multi-lane analysis** - static detectors, compliance, SMT verification, transaction simulation, protocol detection, fuzzing\n- **Validated against professional audit** - 85% coverage on TxPipe's Strike Finance audit findings ([full comparison](AUDIT_COMPARISON.md))\n- **Evidence framework** - findings corroborated across multiple analysis techniques (PatternMatch -> SmtProven -> SimulationConfirmed)\n- **9 output formats** - terminal, JSON, SARIF, Markdown, HTML, PDF, CSV, GitLab SAST, reviewdog\n- **Ecosystem proven** - 10+ real-world projects including SundaeSwap, Anastasia Labs, Strike Finance, and Seedelf (0 crashes)\n\n---\n\n## Quick Start\n\n### Install\n\n```bash\n# Homebrew (macOS/Linux)\nbrew install Bajuzjefe/tap/aikido\n\n# Cargo (Rust >= 1.88.0)\ncargo install --git https://github.com/Bajuzjefe/Aikido-Security-Analysis-Platform aikido-cli\n\n# npm (wrapper)\nnpx aikido-aiken /path/to/project\n\n# Docker\ndocker run --rm -v $(pwd):/project ghcr.io/bajuzjefe/aikido:0.3.1 /project\n\n# From source\ngit clone https://github.com/Bajuzjefe/Aikido-Security-Analysis-Platform.git\ncd aikido && cargo build --release\n```\n\n### Run\n\n```bash\naikido /path/to/your-aiken-project\n```\n\n### Example Output\n\n```\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n AIKIDO v0.3.1 Static Analysis Report\n Project: test/simple-treasury v0.1.0\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\n [CRITICAL] (definite) double-satisfaction - Handler treasury.spend iterates\n outputs without own OutputReference - validators/treasury.ak:23\n\n Spend handler accesses tx.outputs but never uses __own_ref to identify\n its own input. An attacker can satisfy multiple script inputs with a\n single output, draining funds.\n\n 22 | validator treasury {\n > 23 | spend(\n > 24 | datum: Option,\n > 25 | redeemer: TreasuryRedeemer,\n\n Suggestion: Use the OutputReference parameter to correlate outputs\n to this specific input.\n\n ...\n\n 1 critical, 5 high, 7 medium, 0 low, 0 info\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n```\n\n---\n\n## Analysis Architecture\n\nAikido uses a multi-lane approach where independent analysis techniques cross-validate each other:\n\n```mermaid\ngraph TB\n subgraph Input\n A[Aiken Project] --> B[Typed AST]\n end\n\n subgraph Analysis Lanes\n B --> C[Detector Suite
75 detectors]\n B --> D[Compliance
Securify2-style]\n B --> E[SMT Verification
Cardano axioms]\n B --> F[Tx Simulation
exploit generation]\n B --> G[Protocol Detection
DeFi classification]\n B --> H[Fuzz Lane
grammar-aware]\n end\n\n subgraph Evidence Correlation\n C --> I[Evidence Framework]\n D --> I\n E --> I\n F --> I\n G --> I\n H --> I\n end\n\n I --> J[PatternMatch]\n J --> K[PathVerified]\n K --> L[SmtProven]\n L --> M[SimulationConfirmed]\n M --> N[Corroborated]\n\n N --> O[Report
9 output formats]\n\n style C fill:#2EFFB5,color:#0a0a0a\n style D fill:#2EFFB5,color:#0a0a0a\n style E fill:#2EFFB5,color:#0a0a0a\n style F fill:#2EFFB5,color:#0a0a0a\n style G fill:#2EFFB5,color:#0a0a0a\n style H fill:#2EFFB5,color:#0a0a0a\n style I fill:#111111,color:#e0e0e0\n style O fill:#111111,color:#e0e0e0\n```\n\n### Lane Details\n\n| Lane | What it does |\n|------|-------------|\n| **Detector Suite** | Cross-module interprocedural analysis, taint tracking, symbolic execution, delegation-aware suppression, transitive signal merging, datum continuity tracking |\n| **Compliance** | Securify2-style dual-pattern system: every security property has compliance (safe) and violation (unsafe) patterns. 10 security property variants |\n| **SMT Verification** | Solver-independent interface with Cardano domain axioms (value conservation, signature semantics, minting policy). Constraint solving for reachability |\n| **Tx Simulation** | ScriptContext builder generates concrete exploit scenarios. Tests 6 detector categories against simulated transactions |\n| **Protocol Detection** | Automatic DeFi protocol classification (DEX, Lending, Staking, DAO, NFT, Options, Escrow). Token flow and authority analysis |\n| **Fuzz Lane** | Grammar-aware Cardano transaction generation, Echidna-style stateful protocol fuzzing, deterministic PRNG |\n\nSupporting modules: **CWC Registry** (30 entries mapping all 75 detectors), **Scorecard** (Experimental -> Beta -> Stable promotion with quality gates), **SSA IR** (phi nodes, dominators, use-def chains).\n\n---\n\n## Real-World Validation\n\n### Strike Finance Audit Comparison\n\nAikido was benchmarked against TxPipe's professional audit of Strike Finance (perpetuals + forwards contracts):\n\n| Metric | Result |\n|--------|--------|\n| TxPipe security findings analyzed | 24 |\n| Full match (true positive) | 12 |\n| Partial match | 5 |\n| Correctly not flagged (code fixed) | 4 |\n| False negatives | 3 |\n| Aikido unique findings | 26 |\n| **Coverage (unfixed findings)** | **85%** |\n\nFull methodology and per-finding breakdown: **[AUDIT_COMPARISON.md](AUDIT_COMPARISON.md)**\n\n### Ecosystem Validation\n\nValidated against **10+ real-world Aiken smart contract projects** with **zero crashes**:\n\n| Project | Findings | Severity Distribution |\n|---------|----------|-----------------------|\n| SundaeSwap DEX | 47 | 1 critical, 13 high, 24 medium, 3 low, 6 info |\n| Anastasia Design Patterns | 24 | 3 critical, 10 high, 8 medium, 2 low, 1 info |\n| Anastasia Multisig | 6 | 2 critical, 4 medium |\n| Seedelf Wallet | 4 | 1 high, 2 medium, 1 low |\n| Strike Finance (4 repos) | 75 | 5 critical, 18 high, 40 medium, 8 low, 4 info |\n| Acca | 20 | 20 medium |\n| **Total** | **176** | **81% estimated true positive rate** |\n\n---\n\n## Detectors\n\n75 detectors mapped to CWE identifiers.\n\n
\nCritical (5)\n\n| Detector | CWE | Description |\n|----------|-----|-------------|\n| `double-satisfaction` | CWE-362 | Spend handler iterates outputs without referencing own input |\n| `missing-minting-policy-check` | CWE-862 | Mint handler doesn't validate which token names are minted |\n| `missing-utxo-authentication` | CWE-345 | Reference inputs used without authentication |\n| `unrestricted-minting` | CWE-862 | Minting policy with no authorization check at all |\n| `output-address-not-validated` | CWE-20 | Outputs sent to unchecked addresses |\n\n
\n\n
\nHigh (19)\n\n| Detector | CWE | Description |\n|----------|-----|-------------|\n| `missing-redeemer-validation` | CWE-20 | Catch-all redeemer pattern trivially returns True |\n| `missing-signature-check` | CWE-862 | Authority datum fields with no `extra_signatories` check |\n| `unsafe-datum-deconstruction` | CWE-252 | Option datum not safely deconstructed with `expect Some` |\n| `missing-datum-in-script-output` | CWE-404 | Script output without datum attachment (funds locked forever) |\n| `arbitrary-datum-in-output` | CWE-20 | Outputs produced without validating datum correctness |\n| `division-by-zero-risk` | CWE-369 | Division with attacker-controlled denominator |\n| `token-name-not-validated` | CWE-20 | Mint policy checks auth but not token names |\n| `value-not-preserved` | CWE-682 | Spend handler doesn't verify output value >= input value |\n| `unsafe-match-comparison` | CWE-697 | Value compared with match instead of structural equality |\n| `integer-underflow-risk` | CWE-191 | Subtraction on redeemer-controlled values |\n| `quantity-of-double-counting` | CWE-682 | Token quantity checked without isolating input vs output |\n| `state-transition-integrity` | CWE-20 | Redeemer actions without datum transition validation |\n| `withdraw-zero-trick` | CWE-863 | Withdraw handler exploitable with zero-value withdrawal |\n| `other-token-minting` | CWE-862 | Mint policy allows minting tokens beyond intended scope |\n| `unsafe-redeemer-arithmetic` | CWE-682 | Arithmetic on redeemer-tainted values without bounds |\n| `value-preservation-gap` | CWE-682 | Lovelace checked but native assets not preserved |\n| `uncoordinated-multi-validator` | CWE-362 | Multi-handler validator without cross-handler coordination |\n| `missing-burn-verification` | CWE-862 | Token burning without proper verification |\n| `oracle-manipulation-risk` | CWE-20 | Oracle data used without manipulation safeguards |\n\n
\n\n
\nMedium (24)\n\n| Detector | CWE | Description |\n|----------|-----|-------------|\n| `missing-validity-range` | CWE-613 | Time-sensitive datum without validity_range check |\n| `insufficient-staking-control` | CWE-863 | Outputs don't constrain staking credential |\n| `unbounded-datum-size` | CWE-400 | Datum fields with unbounded types (List, ByteArray) |\n| `unbounded-value-size` | CWE-400 | Outputs don't constrain native asset count |\n| `unbounded-list-iteration` | CWE-400 | Direct iteration over raw transaction list fields |\n| `oracle-freshness-not-checked` | CWE-613 | Oracle data used without recency verification |\n| `non-exhaustive-redeemer` | CWE-478 | Redeemer match doesn't cover all constructors |\n| `hardcoded-addresses` | CWE-798 | ByteArray literals matching Cardano address lengths |\n| `utxo-contention-risk` | CWE-400 | Single global UTXO contention pattern |\n| `cheap-spam-vulnerability` | CWE-770 | Validator vulnerable to cheap UTXO spam |\n| `unsafe-partial-pattern` | CWE-252 | Expect pattern on non-Option type that may fail at runtime |\n| `unconstrained-recursion` | CWE-674 | Self-recursive handler without clear termination |\n| `empty-handler-body` | CWE-561 | Handler with no meaningful logic |\n| `unsafe-list-head` | CWE-129 | `list.head()` / `list.at()` without length guard |\n| `missing-datum-field-validation` | CWE-20 | Spend handler accepts datum fields but never validates them |\n| `missing-token-burn` | CWE-862 | Minting policy with no burn handling |\n| `missing-state-update` | CWE-665 | State machine without datum update |\n| `rounding-error-risk` | CWE-682 | Integer division on financial values |\n| `missing-input-credential-check` | CWE-862 | Input iteration without credential check |\n| `duplicate-asset-name-risk` | CWE-694 | Minting without unique asset name enforcement |\n| `fee-calculation-unchecked` | CWE-682 | Fee or protocol payment without validation |\n| `datum-tampering-risk` | CWE-20 | Datum passed through without field-level validation |\n| `missing-protocol-token` | CWE-862 | State transition without protocol token verification |\n| `unbounded-protocol-operations` | CWE-400 | Both input and output lists iterated without bounds |\n\n
\n\n
\nLow / Info (12)\n\n| Detector | Severity | Description |\n|----------|----------|-------------|\n| `reference-script-injection` | Low | Outputs don't constrain reference_script field |\n| `unused-validator-parameter` | Low | Validator parameter never referenced |\n| `fail-only-redeemer-branch` | Low | Redeemer branch that always fails |\n| `missing-min-ada-check` | Info | Script output without minimum ADA check |\n| `dead-code-path` | Low | Unreachable code paths |\n| `redundant-check` | Low | Trivially true conditions |\n| `shadowed-variable` | Info | Handler parameter shadowed by pattern binding |\n| `magic-numbers` | Info | Unexplained numeric literals |\n| `excessive-validator-params` | Info | Too many validator parameters |\n| `unused-import` | Info | Imported module with no function calls |\n\n
\n\nEach detector has detailed documentation with vulnerable examples, safe examples, and remediation guidance. Use `aikido --explain ` or see `docs/detectors/`.\n\n---\n\n## Output Formats\n\n9 formats for different workflows:\n\n```bash\naikido /path/to/project # Colored terminal output (default)\naikido /path/to/project --format json # JSON (machine-readable)\naikido /path/to/project --format sarif # SARIF v2.1.0 (GitHub Code Scanning)\naikido /path/to/project --format markdown # Markdown report\naikido /path/to/project --format html # Standalone HTML report\naikido /path/to/project --format pdf # PDF audit report\naikido /path/to/project --format csv # CSV export\naikido /path/to/project --format gitlab-sast # GitLab SAST\naikido /path/to/project --format rdjson # reviewdog\n```\n\n---\n\n## Configuration\n\n### `.aikido.toml`\n\n```toml\n# Use a preset as a starting point\nextends = \"strict\" # or \"lenient\" for fewer warnings\n\n[detectors]\ndisable = [\"magic-numbers\", \"unused-import\"]\n\n[detectors.severity_override]\nunbounded-datum-size = \"low\"\n\n# Per-file overrides\n[[files]]\npattern = \"tests/**\"\ndisable = [\"hardcoded-addresses\", \"magic-numbers\"]\n```\n\n### Inline Suppression\n\n```rust\n// aikido:ignore[double-satisfaction] -- false positive: own_ref checked in helper\nspend(datum, redeemer, own_ref, tx) {\n ...\n}\n```\n\n### Baseline Support\n\n```bash\naikido /path --accept-baseline # Save current findings as baseline\naikido /path # Only new findings reported\n```\n\n---\n\n## CI/CD Integration\n\n### GitHub Actions\n\n```yaml\nname: Security\non: [push, pull_request]\njobs:\n aikido:\n runs-on: ubuntu-latest\n steps:\n - uses: actions/checkout@v4\n - name: Install Aikido\n run: cargo install --git https://github.com/Bajuzjefe/Aikido-Security-Analysis-Platform aikido-cli\n - name: Run analysis\n run: aikido . --format sarif --fail-on high > results.sarif\n - name: Upload SARIF\n if: always()\n uses: github/codeql-action/upload-sarif@v3\n with:\n sarif_file: results.sarif\n```\n\n### Docker\n\n```bash\ndocker run --rm -v $(pwd):/project ghcr.io/bajuzjefe/aikido:0.3.1 /project --format json\n```\n\n---\n\n## CLI Reference\n\n```bash\naikido /path # Analyze project\naikido /path --fail-on high # Exit non-zero for high+ findings\naikido /path --min-severity medium # Filter to medium+ only\naikido /path --verbose # Include UPLC metrics and budget\naikido /path --list-rules # Show all detectors\naikido /path --explain # Detailed explanation with examples\naikido /path --diff main # Only report findings in changed files\naikido /path --watch # Re-run on file changes\naikido /path -q # Quiet mode\naikido /path --config my.toml # Use specific config file\naikido /path --accept-baseline # Save current findings as baseline\naikido /path --lsp # LSP JSON-RPC diagnostics\naikido /path --interactive # Terminal navigator for findings\naikido /path --fix # Insert suppression comments\naikido /path --generate-config # Generate .aikido.toml from findings\naikido /path --strict-stdlib # Reject stdlib v1.x (default: warn)\naikido --git # Clone and analyze remote repo\naikido --benchmark-manifest benchmarks/local-fixtures.toml --format json\naikido --benchmark-manifest benchmarks/local-fixtures.toml --benchmark-enforce-gates\n```\n\n---\n\n## Architecture\n\n```\naikido/\n├── crates/\n│ ├── aikido-core/ # Library: analysis engine\n│ │ ├── src/\n│ │ │ ├── project.rs # Aiken project loading & compilation\n│ │ │ ├── ast_walker.rs # Typed AST traversal -> ModuleInfo\n│ │ │ ├── body_analysis.rs # Handler body signal extraction + taint tracking\n│ │ │ ├── call_graph.rs # Function dependency graph\n│ │ │ ├── cross_module.rs # Cross-module interprocedural analysis\n│ │ │ ├── symbolic.rs # Symbolic execution & constraints\n│ │ │ ├── delegation.rs # Delegation-aware suppression\n│ │ │ ├── evidence.rs # 5-level evidence framework\n│ │ │ ├── cwc.rs # Cardano Weakness Classification registry\n│ │ │ ├── scorecard.rs # Detector quality tracking & promotion\n│ │ │ ├── ssa.rs # SSA IR with phi nodes & use-def chains\n│ │ │ ├── compliance.rs # Securify2-style compliance analysis\n│ │ │ ├── smt.rs # SMT verification with Cardano axioms\n│ │ │ ├── path_analysis.rs # Path-sensitive analysis & CFG\n│ │ │ ├── invariant_spec.rs # .aikido-invariants.toml DSL\n│ │ │ ├── protocol_patterns.rs # DeFi protocol detection & classification\n│ │ │ ├── tx_simulation.rs # ScriptContext builder & exploit generation\n│ │ │ ├── fuzz_lane.rs # Grammar-aware tx fuzzing\n│ │ │ ├── config.rs # .aikido.toml configuration\n│ │ │ ├── suppression.rs # Inline suppression comments\n│ │ │ ├── baseline.rs # Baseline file support\n│ │ │ ├── uplc_analysis.rs # UPLC bytecode analysis & budgets\n│ │ │ └── detector/ # 75 detector implementations\n│ │ └── tests/ # 1186+ tests\n│ └── aikido-cli/ # Binary: clap-based CLI\n├── fixtures/ # Test contracts (7 fixtures)\n├── audits/ # Audit comparison artifacts\n├── docs/detectors/ # Per-detector documentation\n├── .github/workflows/ # CI, cross-platform, release, fuzz\n├── homebrew/ # Homebrew formula\n├── npm/ # npm wrapper package\n├── vscode-extension/ # VS Code extension\n├── fuzz/ # cargo-fuzz targets\n└── Dockerfile # Multi-stage build\n```\n\n### Building from Source\n\n```bash\ngit clone https://github.com/Bajuzjefe/Aikido-Security-Analysis-Platform.git\ncd aikido\ncargo build --release # Binary at target/release/aikido\ncargo test # Run test suite (1186+ tests)\ncargo clippy --all-targets # Lint (zero warnings)\n```\n\nRequires Rust >= 1.88.0.\n\n---\n\n## Vulnerability Coverage\n\nDetectors are derived from real vulnerabilities found in published Cardano smart contract audits:\n\n| Source | Findings Covered |\n|--------|-----------------|\n| MLabs audit reports | Double satisfaction, missing minting policy, arbitrary datum, unbounded size |\n| Vacuumlabs audit reports | Unbounded value size, token dust attacks |\n| [Plutonomicon](https://github.com/Plutonomicon/plutonomicon) | Unrestricted minting, double satisfaction patterns |\n| Anastasia Labs audit reports | Staking credential theft, datum handling |\n| TxPipe audit reports | Oracle manipulation, state transition integrity, value preservation |\n| CWE Database | 75 detectors mapped to specific CWE identifiers |\n\n---\n\n## License\n\nMIT\n\n\n", "bytes": 20383, "sha": "f89245934c2c741ebe0008fb00a564b27f473ff877a218820c96bf75be83179b", "repo_slug": "bajuzjefe/aikido-security-analysis-platform", "fonte": "repo", "truncated": false, "api": "https://agentalog.com/api/listings/mcp_io_github_bajuzjefe_aikido_mcp_d3c08397/readme" }