Roslyn-Backed-MCP: Instructions file for Codex

AGENTS Bootstrap

Use this file as the stable AI bootstrap entry point for this repository.

File Purpose (Critical)

This file is a bootstrap router, not a complete instruction set. Always execute Session Start (Required) before performing any task. Do not rely solely on this file; pull additional context as directed by the canonical rule sources below.

Standing Engineering Directives

Restated from ~/.claude/CLAUDE.md (canonical source). These eight directive cores (the bold titles) override expedience and are verbatim — do not summarize, drop, or alter them. The one-line gloss after each is a condensed summary for quick reference; the authoritative Fires/Prevents/Edge detail lives in ~/.claude/CLAUDE.md.

1. Correct fix > quick fix. When a quick fix and a correct fix are both viable, choose the correct fix. The correct fix addresses the root cause; the quick fix patches a symptom. Quick fixes are only acceptable when the correct fix is genuinely out of scope for the current task — and when that happens, file a backlog row (per rule #3) for the correct fix before shipping the quick one. "We'll fix it later" without a tracked row = does not exist.

2. Optimize for AI consumption by default. 99% of files in this repo are AI-facing: AGENTS.md, CLAUDE.md, .github/copilot-instructions.md, .github/prompts/**, ai_docs/**, planning docs, runtime references, audit reports. Optimize them for fast, dense parsing: tables over prose, structured data over paragraphs, pointers over duplication, short imperative sentences over narrative. Human-facing files (README.md at landing-page level, docs/**) get human prose. Everywhere else, AI-optimized.

3. Bad code is never silent — this rule fires in EVERY coding session. Not just doc-audit. Not just code review. Every session where you read, edit, fix, refactor, debug, navigate, or otherwise touch code. When you observe bad code — dead code paths, swallowed exceptions, hardcoded values that should be config, hardcoded credentials/secrets, commented-out blocks, TODO/FIXME/HACK/XXX markers without a tracking row, obvious anti-patterns, security smells, broken naming, suspiciously stale comments, copy-paste duplication, god classes, layering violations, sync-over-async, untested critical paths — you MUST (a) call it out in your response and (b) recommend an appropriately-sized backlog row (≤4 production files, ≤3 test files, one regression shape). Applies whether the bad code is the file you were sent to edit, an adjacent file you opened for context, an import target, a test file, or anything you read in the course of solving the task. Critically: if you are FIXING a code section and it is bad/poorly-written/incorrect, surfacing it is mandatory even if the fix is "in scope" — the act of editing does not absolve the obligation to flag. The right time to fix the bad code is when the row's priority demands it, not when you happen to be in the file — but the row MUST exist. Silently editing around bad code, polishing a bad pattern without flagging, or "I'll mention it if asked" all count as misses.