LOADING PAGE
Preparing your page.
Your site data stays in this session while the page loads.
AGENT INDEX
Coding Agents & Repository Guidelines Library
Best practices and operational guides for structuring repo rules, prompt boundaries, testing conventions, and prohibited patterns for Cursor, Claude Code, GitHub Copilot, and agentic workflows.
25 technical guides
IDEs Cursor · Copilot · Claude · Windsurf
Static pre-rendered pages
All Agent Guidelines
25 published articlesAGENTS.md Best Practices: Clear Instructions for Coding AgentsArchitectural blueprint for authoring production AGENTS.md manifests: hierarchical inheritance, hard negative invariants, deterministic build gates, and multi-agent coordination.broadly supportedAGENTS.md vs .cursorrules: Open Multi-Agent Standard vs Vendor-Locked IDE RulesArchitectural comparison between repository-level open specification AGENTS.md and vendor-specific .cursorrules: portability, scoping, and multi-IDE alignment.broadly supportedAI-Readable Changelogs: Make Releases Easy to VerifyArchitectural guide for structuring machine-readable changelogs: Keep a Changelog standards, SemVer guarantees, migration diffs, and agentic dependency upgrades.broadly supportedAutomating llms.txt: Generate and Validate It in CIProduction guide for automating llms.txt generation: GitHub Actions CI/CD workflows, dead link validation, token budgeting gates, and edge CDN cache purges.broadly supportedAutonomous Agents vs Inline Copilots: Execution vs AutocompleteArchitectural comparison between autonomous CLI coding agents and inline copilots: security boundaries, rule enforcement, token budgets, and test-fix loops.broadly supportedClaude Code Optimization: Organize CLAUDE.md Without Wasting ContextProduction guide for optimizing Anthropic's Claude Code CLI: hierarchical CLAUDE.md architecture, pre-bash safety hooks, prompt caching, and cost engineering.broadly supportedClaude Code: reviewing machine contextMasterclass engineering guide for operating Anthropic's Claude Code CLI: context budgeting, prompt caching, tool loop governance, and automated verification.broadly supportedCLAUDE.md vs AGENTS.md: Terminal CLI Subagent Memory vs Repository GovernanceArchitectural differentiation between Claude Code CLI memory protocols in CLAUDE.md and repository-wide governance manifests in AGENTS.md.broadly supportedCodebase Context Strategies: Give AI Agents the Right Repository MapArchitectural strategies for navigating enterprise codebases: Tree-sitter AST symbol graphs, PageRank repo maps, semantic search, and dynamic LSP context.broadly supportedContext Injection Techniques: How IDE Agents Select Repository FilesArchitectural analysis of IDE context injection pipelines: Jaccard tab similarity, LSP symbol resolution, static instruction precedence, and context truncation algorithms.broadly supportedCursor IDE Optimization: Scope Repository Context Without NoiseAdvanced engineering guide for optimizing Cursor IDE: .cursorignore tuning, modular .mdc rules, Merkle tree codebase indexing, and Composer agent controls.broadly supportedCursor: using site context carefullyProduction guide for optimizing Cursor IDE: indexing external docs via @Docs, configuring modular .cursor/rules, and executing safe agentic refactoring.broadly supportedDocumenting Architecture for AI: Describe Boundaries and DecisionsArchitectural guide for structuring system documentation for autonomous AI agents: C4 models, Architecture Decision Records (ADRs), and invariant boundaries.broadly supportedGitHub Copilot Context: Which Repository Instructions Does It Read?Architectural guide to GitHub Copilot's context ingestion: .github/copilot-instructions.md, scoped .github/instructions/*.md, and AGENTS.md inheritance.broadly supportedIndirect Prompt Injection in AI Context: Poisoning vs HardeningThreat modeling indirect prompt injection vectors in autonomous agent context pipelines: poisoned documentation, image callbacks, and AGENTS.md defenses.broadly supportedMCP Tools vs llms.txt: Dynamic Tool Calling vs Static ContextArchitectural comparison between Anthropic Model Context Protocol (MCP) servers and static llms.txt endpoints: latency, compute overhead, and reliability.broadly supportedMulti-Agent Context Handoffs: Massive Shared Context vs Minimal State PacketsArchitectural comparison between shared-context multi-agent systems and partitioned minimal packet handoffs across Planner, Coder, and QA agents.broadly supportedProhibitions in AGENTS.md: Tell Agents What Not to DoArchitectural guide for authoring non-negotiable prohibitions in AGENTS.md: pairing negative rules with mandatory alternatives and CI enforcement.broadly supportedPrompt Engineering in Code: Version the Behavior You ShipProduction guide for engineering prompts inside application code: strongly-typed template builders, Zod schema validation, few-shot management, and CI eval suites.broadly supportedRoot AGENTS.md vs Nested Rules: Monorepo Scoping ArchitectureArchitectural comparison between single repository root AGENTS.md manifests and hierarchical nested directory rules in complex polyglot monorepos.broadly supportedStructuring Monorepos for AI: Scope Context by PackageArchitectural strategies for navigating enterprise monorepos with AI agents: package boundaries, dependency graphs, scoped AGENTS.md manifests, and Turborepo filters.broadly supportedSystem Prompts in a Repository: Version the Contract, Not the SecretEngineering guide to managing system prompts in version control: GitOps workflows, prompt regression testing, secret sanitization, and indirect prompt injection defense.broadly supportedTesting Conventions for Agents: Make AGENTS.md VerifiableArchitectural guide for establishing verifiable testing conventions for AI agents: test-first loops, fast feedback tiers, fixture isolation, and anti-cheating guards.broadly supportedVersioning AI Context: Keep Guidance Aligned With CodeArchitectural blueprint for versioning AI context files: SHA-256 provenance tracking, SemVer for agent instructions, edge cache synchronization, and release tag alignment.broadly supportedWindsurf Cascade vs Cursor Composer: Context Memory ArchitectureArchitectural comparison between Codeium Windsurf Cascade and Cursor Composer: codebase AST indexing, real-time tool calling, and context retention models.broadly supported
Generate an AGENTS.md for your repository.
Scaffold stack presets, architecture rules, testing commands, and agent prohibitions in the studio.