If you discover a security vulnerability in Crashpad, please report it responsibly.
DO NOT open a public GitHub issue for security vulnerabilities.
Instead, email: [email protected] (placeholder — replace with actual contact)
- Description of the vulnerability
- Steps to reproduce
- Potential impact
- Suggested fix (if any)
- Acknowledgment: Within 48 hours
- Initial assessment: Within 1 week
- Fix & release: Depends on severity (critical issues within 72 hours)
Crashpad is a safety-critical component. Keep in mind:
-
RPC Endpoint Trust: Crashpad simulates via your RPC endpoint. If the endpoint is compromised, simulations may be unreliable. Use trusted, dedicated RPC providers.
-
Rules Configuration: The safety rules YAML file is the source of truth for limits. Ensure it's:
- Version controlled
- Reviewed before changes
- Not writable by the agent process itself
-
Private Key Separation: Crashpad never handles private keys. Transaction signing should happen in a separate, hardened process.
-
Simulation ≠ Guarantee: Fork simulation approximates on-chain behavior but cannot account for:
- Mempool state changes between simulation and inclusion
- Block-level reorgs
- Oracle manipulation at the exact block of inclusion
-
MEV Protection: Private routing reduces but does not eliminate MEV risk. Relay operators are trusted parties.
The following are in scope for security reports:
- Bypass of safety checks (slippage, position limits, loss caps)
- Injection via YAML rules files
- Crash or denial of service via crafted transaction input
- Incorrect price impact or slippage calculations
- Reward-hack detection evasion
- Social engineering attacks
- Compromised RPC endpoints (report to the provider)
- Issues in upstream dependencies (report to them directly)