Ethereum Classic has just been through a live lesson in what happens when a chain's official-looking accounts outrank its actual maintainers. A disputed Core Geth v1.13.0 release, promoted as a security update by the @ETC_Network account on September 15, reached several mining pool nodes before operators recognized the problem and reversed the migration. By September 16, Classix — the group behind the maintained Core Geth line — had published an incident report calling the release rogue, urging operators to stay on Argos v1.12.23 and asking GitHub administrators to lock down the ethereumclassic organization's repositories. No blocks were lost, no chain reorganizations occurred and no funds were affected. The classification was high severity, low impact — because the near-miss is the story.
The backstory is a version-control fight wearing a security advisory's clothes. The ethereumclassic/core-geth repository was forked from etclabscore/core-geth in December 2024, then went quiet — until September 12, when a v1.13.0-rc1 tag appeared. Six release candidates followed in two days, with 96 commits containing 13,422 added and 3,977 deleted lines pushed directly to the main branch in 56 hours, with no pull requests and no outside review, before v1.13.0 was labeled stable at 15:06 UTC on September 14. The next morning, @ETC_Network told node operators to migrate, similar messaging appeared on CoinMarketCap, and mining pools received emails from an ethereumclassic.com address.
What Core Geth v1.13.0 Actually Changed
The release told operators that every v1.12.x node carried unpatched vulnerabilities and should upgrade. Classix reviewed the seven cited security issues and found five had already been fixed in the maintained line between March and August — CVE-2026-22862 and CVE-2026-26315 in Aegis v1.12.21, cryptographic issues in Hermes v1.12.22, and delayed peer-to-peer message decoding in Argos v1.12.23 — while the other two did not touch ETC's peer-to-peer path. One, a KZG-proof issue, applies to blob transactions from Ethereum's Cancun upgrade, which ETC never activated. Core Geth maintainer Diego López León reviewed the remaining differences and found no exploitable flaw in Argos that v1.13.0 corrected.
The real changes were in consensus behavior and network infrastructure. One commit reenabled Modified Exponential Subjective Scoring — MESS, the anti-reorg mechanism ETC adopted in 2020 and deliberately disabled through ECIP-1110 after Ethereum moved to proof of stake — by removing the configuration that turned it off at block 19,250,000. Classix warned that if only Core Geth nodes run MESS while Besu, Nethermind and Getc do not implement it, the network could see divergent chain selection across clients on a chain with a painful reorg history, including the three majority attacks of August 2020. Another commit replaced a DNS tree signing key maintained by etclabscore contributors since 2020, hardcoded three new bootnode IP addresses, removed the two established discovery trees, and hosted three replacement domains on a single Cloudflare account — an obvious single point of failure the repository itself acknowledged. Operators were never told who controls the new signing key.
How It Ended — and Why It Almost Didn't
Node status data cited by Classix shows four 2Miners nodes running v1.13.0 at 12:09 UTC on September 15; by 23:35 UTC all four had returned to Argos v1.12.23. Other pools never moved. Etcnodes.org counted 11 v1.13.0 nodes at 07:33 UTC on September 15, down to 10 by the next day, three of which matched the bootnode IPs hardcoded into the new client. The migration guide had also instructed operators to rotate their P2P node keys over a vulnerability Classix says was fixed in March — a rotation that changes a node's network identity and forces it to rebuild peer connections through, conveniently, the new discovery infrastructure.
The trust mechanics deserve the attention here. Ethereum Classic's own website disclaims any official developer, maintainer, website or client — the network's legitimacy rests on public history and adoption, not branding. That vacuum is what the ethereumclassic.com address and @ETC_Network account exploited, however benign their holders' intent may have been. It is the supply-chain version of a problem crypto users know in another form — fake infrastructure convincing people to run the wrong code — and it rhymes with the review-and-consensus fights other networks are having about who may ship software at all, a fight Bitcoin's own community is conducting in slow motion.
What Classix Wants Next
The report recommends operators avoid v1.13.0, restore rotated node keys, check MESS configuration, and consider client diversity — running Nethermind, Besu or Getc rather than concentrating hash power on one codebase. It asks the ethereumclassic GitHub organization to require proposals and reviews before new repositories are created, protect default branches, identify maintainers, and archive or delist core-geth with a warning that it is not an official client. Classix calls the document an interim report that will be updated as remaining nodes are verified and maintainers respond. The takeaway for every chain without a designated maintainer is uncomfortable: anyone can fork a repository, grab a lookalike domain and an account name, and start emailing mining pools. On ETC, the guarded process caught it in a day. The next chain may not be as lucky, and the honest question the incident leaves is whether that is a governance flaw or simply what decentralization looks like at 3 AM.












































