Multi-agent (LangGraph + Claude) app that turns an earnings call into a source-attributed analyst brief: ingest - FinBERT tone + KPI-vs-consensus - SurpriseSignal - grounding check - delivery.
15 files for Claude Code, GitHub Copilot, Codex and OpenCode: security-reviewer, test-writer, /eval, /new-agent and 11 more — 14,546 tokens loaded in every session.
.github/copilot-instructions.md A 4,584 tok .github/instructions/evals.instructions.md A 979 tok .github/instructions/prompts.instructions.md A 639 tok .github/instructions/python.instructions.md A 594 tok AGENTS.md A 4,591 tok CLAUDE.md A 3,159 tok .claude/settings.json A — .claude/skills/prompt-linter/SKILL.md A 17 tok .claude/commands/eval.md A 14 tok .claude/commands/new-agent.md A 23 tok .claude/commands/review-pr.md A 18 tok .github/prompts/add-skill.prompt.md A 18 tok .github/prompts/add-tool.prompt.md A 32 tok .claude/agents/security-reviewer.md A 15 tok .claude/agents/test-writer.md A 15 tok These files are suhaas/earnings-edge's own configuration — they tell Claude Code, GitHub Copilot, Codex and OpenCode how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.