AntFleet

Disagreement · 9bb17981-anthropic-2

Trusted-author set inconsistent with documented precedence (aeonframework vs aaronjmars)

solo Opus
repo 6f7fc663·PR #7·reviewed 1 week ago

Opus finding

Trusted-author set inconsistent with documented precedence (aeonframework vs aaronjmars)

lowdocs-gapmedium
  • skills/pr-triage/SKILL.md:15-22
Two hardcoded trusted logins are baked into the skill prose. If the fleet operator's login is neither (a fork user), every PR — including the operator's own — gets triaged as external. The doc reads 'the agent's own login' but then lists two specific humans. For a forkable fleet skill this should be parameterized (e.g., read from MEMORY.md or aeon.yml), otherwise downstream forks inherit triage of their own PRs as 'external'.

Recommendation

Resolve the agent's own login dynamically (`gh api user --jq .login`) or from `memory/MEMORY.md`/`aeon.yml`, rather than hardcoding `aeonframework` / `aaronjmars`.

Other reviewer

The other reviewer flagged nothing in this file/line range.

Why this didn't post

This finding didn't meet AntFleet's unanimous agreement threshold. Both frontier models review every PR independently; only findings they both flag with the same severity and category are posted to the PR. This one fell through.

read the methodology →