feat: Harbor framework integration for Terminal-Bench evaluations#202

Merged

jeremyeder merged 22 commits intoambient-code:mainfrom

jeremyeder:002-harbor-real-integration

Dec 10, 2025

Commits on Dec 5, 2025

chore: update leaderboard data [skip ci]
github-actions[bot]
committed

Commits on Dec 8, 2025

fix: resolve 45 test failures across CLI, services, and assessors (#4)

jeremyeder
and
claude
authored
chore(release): 2.10.0 [skip ci]
semantic-release-bot
committed

Commits on Dec 9, 2025

fix: resolve 45 test failures across CLI, services, and assessors (#4)

jeremyeder
and
claude
committed
chore(release): 2.10.0 [skip ci]

semantic-release-bot
authored and
jeremyeder
committed
feat: add Harbor framework integration for real Terminal-Bench evaluations

jeremyeder
and
claude
committed
feat: implement blocking test strategy with tiered CI jobs

jeremyeder
and
claude
committed
fix(security): implement critical security fixes from code review

jeremyeder
and
claude
committed
fix: correct Harbor results parsing to match actual Harbor 2.0 JSON structure

jeremyeder
and
claude
committed
chore: save Harbor integration WIP before rebase onto v2.15.0

jeremyeder
and
claude
committed
chore: restore version to 2.15.0 after rebase
jeremyeder
committed
fix: remove duplicate assessor registration for architecture_decisions and issue_pr_templates
jeremyeder
committed
feat: redesign assess command output with detailed results table

jeremyeder
and
claude
committed
fix: validate API key before HarborConfig initialization
jeremyeder
committed
feat: add automatic Harbor CLI preflight checks with dataset management

jeremyeder
and
claude
committed
Merge branch 'main' into 002-harbor-real-integration
jeremyeder
committed
fix: pass full environment to Harbor subprocess

jeremyeder
and
claude
committed
fix: set ANTHROPIC_AUTH_TOKEN for Harbor's Claude Code agent

jeremyeder
and
claude
committed