Dev frustrated building agent harness with Claude Code, asks for best practices
md1630 · reddit · 2026-10-15
A developer whose Claude Code-built agent (with multiple skills) has been live for weeks complains it's inflexible: it misreads intent, skips the right skill, uses crude top-level regex rules instead of interpreting intent, and treats 'search internally and the web' as a mandate to always do both—even erroring on web searches when the answer was internal, without returning links. Concluding that Claude Code itself doesn't know harness engineering best practices, he plans to study tutorials and design the harness himself, and asks Reddit for recommended guides.
More from coding & agent
- Index Network open-sources Social Discovery Protocol: agents negotiate fit on your behalf — carsonfarmer · 2026-10-23
- Installed a formal-verification skill: it silently installed Java, MacBook runs 10 degrees hotter — zealcaiden · 2026-10-23
- OpenAI's Decisions API hits GA: near real-time routing up to 10x faster than GPT-6 Luna — OpenAIDevs · 2026-10-23
- Ruff author warns: upgrade Node/Bun now — --use-system-ca still accepts the eMudhra cert — charliermarsh · 2026-10-23
- Proof of Agent: paste one prompt and your AI mines, grades, and self-governs — adamamcbride · 2026-10-23
- DeepLearning.AI weekly: GPT-6.1 hits 97.9% byte parity on COBOL migration — DeepLearningAI · 2026-10-23