Pyde is a Layer 1 blockchain in pre-mainnet development.
This policy applies to every repo in the
pyde-net org unless an individual repo
overrides it with a more specific SECURITY.md.
Do not open a public issue for security vulnerabilities.
Email: security@pyde.network (or until that's live: info@pyde.network,
the project contact via README)
Please include:
- A description of the vulnerability
- Steps to reproduce (or a proof-of-concept if you have one)
- Which repo / commit / module is affected
- Your assessment of the severity and impact
- Whether the issue is exploitable in the current code or only after a hypothetical future change
We will respond within 5 business days acknowledging receipt. A full triage decision typically follows within 14 days.
In scope:
- All public code in
pyde-netorg repos - Cryptographic implementations in
pyde-crypto - The WASM execution layer (post-pivot, to be re-cut in a fresh engine workspace; the pre-pivot reference is in
archive) - The retired Otigen compiler in
otic, a historical artifact, though logic bugs that could resurface post-pivot are still in scope - Wire-format / consensus / mempool design issues
- Logic bugs in
pyde-bookcompanion specs that, if implemented as documented, would create a vulnerability
Out of scope:
- The pre-pivot codebase in
archive(archived, read-only, not part of the active project) - Third-party dependencies (report those upstream, and we'll happily help with disclosure coordination, but the patch belongs there)
- Issues in private/internal infrastructure (CI keys, secrets, etc.); report directly via email regardless of "scope"
- Social-engineering issues unrelated to protocol design
- Issues that require a compromised wallet, browser, or operator endpoint (i.e., classes that any chain has)
We use a standard four-level scale:
| Severity | Examples |
|---|---|
| Critical | Funds theft, consensus break, double-spend, key extraction, signature forgery |
| High | DoS that halts the chain or a validator, privilege escalation, MEV reintroduction |
| Medium | Significant correctness bugs without funds-loss path, partial DoS |
| Low | Edge-case correctness issues, hardening gaps, documentation issues with security implications |
Standard responsible-disclosure with reasonable flexibility:
- Day 0: Report received, acknowledged within 5 business days
- Day 14 to 90: Triage, fix, internal review. We'll keep you updated
- Disclosure: Coordinated public disclosure once a patch is released, or after 90 days (whichever is sooner), unless we explicitly agree on an extended embargo for serious issues
We will credit you in the disclosure unless you prefer anonymity.
We will not pursue legal action against good-faith security research that:
- Complies with this policy
- Avoids privacy violations, destruction of data, or service interruption
- Reports findings promptly via the channel above
- Does not exploit a finding beyond what's needed to demonstrate it
A formal bug bounty program will launch with the incentivized testnet (Phase 9 of the launch plan). Pre-mainnet vulnerability reports are not currently compensated financially but will be credited at the bounty program's launch, with priority consideration for retroactive recognition.
Pyde is pre-launch. The codebase is being prepared for external audit
across five specialist tracks (consensus, WASM execution layer, cryptography,
networking, otigen toolchain) per
pyde-book Chapter 19.
Vulnerabilities found in the pre-audit code are valuable input to the
audit, not a sign the audit has failed. We welcome them.
Specifically, the multi-region performance harness build-out (specified in pyde-book companion/PERFORMANCE_HARNESS.md) is the gating prerequisite before benchmark numbers are published, and the work that surfaces it tends to surface vulnerability classes too.
One known constraint surfaces here, tracked for post-mainnet hardening:
ml-kem(Kyber-768, used for transport-layer encryption) is at0.3.0-rc.0(NIST FIPS 203 release candidate); upgrading to the stable release is on the post-mainnet checklist.
Reports against this known constraint are appreciated as confirmation or expansion of impact, but the constraint itself is not novel.
This security policy is licensed under CC0, so copy and adapt freely.