|
All checks were successful
CI / validate (push) Successful in 1m21s
Per saved memory feedback_act_runner_needs_config_file_env: runners 1-8 were spawned without -e CONFIG_FILE=/config.yaml; act_runner fell back to /data/config.yaml and ignored runner.envs the whole time. Orchestrator recreated 1-8 with full proper env. All 16 now uniform with AGENT_TOOLSDIRECTORY + RUNNER_TOOL_CACHE + GITHUB_SERVER_URL + GH_HOST. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|---|---|---|
| .github/workflows | ||
| .molecule-ci/scripts | ||
| adapters | ||
| runbooks | ||
| skills/cross-vendor-review | ||
| .gitignore | ||
| CLAUDE.md | ||
| known-issues.md | ||
| plugin.yaml | ||
| README.md | ||
molecule-skill-cross-vendor-review — Adversarial Multi-Model Review
Plugin for Claude Code. Run an adversarial code review against a non-Claude model (Codex, GPT, or Gemini) and surface disagreements with Claude's own review. Two LLMs catch bugs one doesn't.
Use ONLY for high-stakes PRs: auth, billing, data-deletion, irreversible migrations, large blast-radius changes.
When to use
Install on workspaces that review high-stakes PRs. The skill is invoked on-demand
via the Skill tool after Claude has completed its own review.
What it does
- Presents the PR diff to a second-model reviewer (configurable: Codex, GPT, Gemini)
- Asks the second model to flag concerns independently
- Surfaces disagreements with Claude's own review
- Highlights where Claude may have missed something the second model caught
Installation
In org template (org.yaml)
plugins:
- molecule-skill-cross-vendor-review
From URL
github://Molecule-AI/molecule-ai-plugin-molecule-skill-cross-vendor-review
License
Business Source License 1.1 — © Molecule AI.