{"slug": "two-tiny-claude-code-skills-that-fixed-my-two-biggest-agent-problems", "title": "Two tiny Claude Code skills that fixed my two biggest agent problems", "summary": "A developer at dualform labs released two open-source skills for Claude Code that address common agent failures: spec, which forces upfront clarification of ambiguous tasks, and review-audit, which performs a read-only audit across six axes with evidence. Both are single prompt files under Apache-2.0 with no dependencies.", "body_md": "Two open-source skills for Claude Code. Each is a single prompt file, Apache-2.0, no dependencies. Repos at the bottom.\n\nWorking with a coding agent, I kept hitting the same two failure modes. Not \"the model can't write code\" — it writes code fine. The failures were upstream and downstream of the code: **the agent guessing on an ambiguous task**, and **me trusting a review that hadn't actually checked anything.**\n\nSo I built one small skill for each. Here's what they do and why they're shaped the way they are.\n\nHand a vague task to an agent and you watch the same thing happen. It guesses. It drifts. It quietly makes a call you'd have made differently — and you find out after the code is written, when changing your mind is expensive. The cost isn't the typing. It's the rework.\n\n** spec** moves the decisions to the front. You run\n\n`/spec <one-line idea>`\n\n, it reads your repo and the conversation, then asks only what it Two things keep it honest:\n\n`done`\n\nonly after it ran on a real case and the output was shown.The whole skill is one prompt file (`SKILL.md`\n\n). No build, no dependencies.\n\nAsk an AI to \"review this change\" and you can get a confident, plausible **PASS** — that skipped half the checks, cited no evidence, and never ran the tests. A green light you can't trust is worse than no review.\n\n** review-audit** is a read-only, single-pass audit over your change across six axes: correctness, wiring (built-but-never-called / dead code), security, test efficacy, spec compliance, and regression. The discipline is simple and strict:\n\n`file:line`\n\n+ grep/run evidence. \"I didn't check this\" is a first-class output, not a silent gap.`file:line`\n\n. \"Looks fine\" isn't evidence.It runs in the calling agent's own context — no sub-agent fan-out — so it stays cheap enough to run on every change. When one pass genuinely isn't enough (a release gate, a high-risk change), it tells you, in its own output, to escalate.\n\nBoth are one prompt file each.\n\n```\n# spec\ngit clone https://github.com/dualform-labs/spec-skill.git\ncp -r spec-skill/skills/spec ~/.claude/skills/\n\n# review-audit\ngit clone https://github.com/dualform-labs/review-audit.git\ncp -r review-audit/skills/review-audit ~/.claude/skills/\n```\n\nThen in Claude Code: `/spec a menu-bar app that warns me when my Mac is thermally throttled`\n\n, or `/review-audit`\n\non a change before you call it done. Output language is `auto`\n\n/ `ja`\n\n/ `en`\n\n.\n\nNo network calls (Claude Code only), no telemetry, no bypass-permissions.\n\nThese are prompt-file skills, not magic. Single-pass detection in `review-audit`\n\nis model-dependent; if you need per-run proof of detection power or fresh-context adversarial verification, that's the heavier `review-audit-pro`\n\ntier (coming soon). And `spec`\n\nwon't make a bad idea good — it just makes the decisions explicit before code is written, where they're cheap to change.\n\nIf you try them, I'd genuinely like to hear where they break or annoy you.\n\n— a dualform project", "url": "https://wpnews.pro/news/two-tiny-claude-code-skills-that-fixed-my-two-biggest-agent-problems", "canonical_source": "https://dev.to/dualform/two-tiny-claude-code-skills-that-fixed-my-two-biggest-agent-problems-13fh", "published_at": "2026-06-14 03:57:16+00:00", "updated_at": "2026-06-14 04:28:40.673316+00:00", "lang": "en", "topics": ["developer-tools", "ai-agents", "large-language-models"], "entities": ["dualform labs", "Claude Code", "spec", "review-audit"], "alternates": {"html": "https://wpnews.pro/news/two-tiny-claude-code-skills-that-fixed-my-two-biggest-agent-problems", "markdown": "https://wpnews.pro/news/two-tiny-claude-code-skills-that-fixed-my-two-biggest-agent-problems.md", "text": "https://wpnews.pro/news/two-tiny-claude-code-skills-that-fixed-my-two-biggest-agent-problems.txt", "jsonld": "https://wpnews.pro/news/two-tiny-claude-code-skills-that-fixed-my-two-biggest-agent-problems.jsonld"}}