feat: AVE-2026-00074 -- reclaimable dead external anchor (SkillJacking) - #135
Merged
Conversation
…ing) A skill referencing an external GitHub owner, package, domain, or cloud subdomain that was live when authored can have that anchor go dead and become claimable by an attacker, with no change to the skill's own content. Distinct from AVE-2026-00062 (absence of pinning at declaration time): this class is a previously-valid reference decaying after the fact, not a reference that was never pinned. Sourced from repo-forensics' scan_dead_anchors.py and AIR's SkillJacking disclosure (925 skills / ~134,000 agents on hijackable dependencies, including a confirmed takeover of an 11,483-install skill via GitHub owner re-registration). Co-Authored-By: Claude <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
provenance_vectoragainst AVE-2026-00062 (genuinely distinct mechanism), and by reading the actual repo-forensics scanner source, not just its README:scan_dead_anchors.py-- extracts every external anchor a skill points at and probes whether it's confirmed-claimable, live-and-owned, or unverifiable.seedance2-apiskill (11,483 installs) via re-registering its deleted GitHub owner.cvss_base8.7,aars5.5 --dynamic_identityscored at genuine maximum, this class is definitionally trust-anchor confusion).mitre_atlasresearched and left empty rather than force-fit: ATLAS's own AI-software supply-chain technique (AML.T0010.001) names build/maintainer compromise, namesquatting, and hallucinated package names as its sub-cases -- none name an attacker legitimately re-registering an identity the original owner abandoned. A genuine gap in ATLAS's own taxonomy.Test plan
python3 scripts/validate_records.py-- 74/74 records validpython3 scripts/check_fixtures.py-- all records have positive + negative fixturespytest tests/ -x -q-- 297 passednode scripts/build-records.js-- dist regenerated, frozen v1.1.0 snapshot untouched