Skip to main content

Posts

Featured

New ask Hacker News story: Ask HN: How do you give AI agents real codebase context without burning tokens?

Ask HN: How do you give AI agents real codebase context without burning tokens? 4 by donhardman | 0 comments on Hacker News. Working on a large Rust codebase. The token problem is real — Claude Code will happily spend $5 of context just trying to understand how two modules relate before writing a single line. And once context compaction kicks in, it's even worse — the agent loses the thread completely and starts grepping the same files again from scratch. Approaches I've tried: Feeding CLAUDE.md / architecture docs manually — helps, but gets stale fast. Cursor's built-in indexing — breaks on monorepos, and I don't love proprietary code going to their servers. Basic MCP server with grep — works for exact matches, useless for semantic queries. Eventually built something more serious: a local Tree-sitter indexer that builds a knowledge graph of file relationships and exposes it via MCP so agents query semantically instead of grepping blind. One tool call instead of 15 g...

Latest Posts

New ask Hacker News story: Amazon degraded shopping- you have to put in cart to see the price

New ask Hacker News story: HATEOAS Works with an LLM in the Mix

New ask Hacker News story: Stathat Is Shutting Down

New ask Hacker News story: Ask HN: How do you find contracting/freelance roles without recruiters nowadays?

New ask Hacker News story: Ask HN: Maintainers, do LLM-only users often clutter your issues/PRs?

New ask Hacker News story: Ask HN: What's your experience self-hosting in 2026?

New ask Hacker News story: Ask HN: Best use / examples of agents / OpenClaw that you saw recently?

New ask Hacker News story: Tell HN: GitHub Having Issues

New ask Hacker News story: Google violates its 14-day deprecation policy for Gemini 3 Pro Preview

New ask Hacker News story: Google just killed my project