v6.4.2
debpalash/VoiceStudiov6.4.2Sep 25, 2026by obra
Upgrading?
AI Summary
The `writing-plans` skill has been optimized to be faster and leaner by recording implementation decisions (signatures, test assertions) rather than full code, preventing overzealous implementation and reducing token usage.
Key Highlights
- Optimized `writing-plans` skill produces leaner plans and executes faster
- Plans now record decisions like signatures and test assertions instead of writing full code
- Self-review checks plan length proportion to the spec to prevent transcripts
- Plans verified against planted-defect probes with 100% success rate
Breaking Changes
- Removed `plan-document-reviewer-prompt.md` file
- Removed `CLAUDE.md` (replaced by `AGENTS.md`)
- Plan structure changed from 'No Placeholders' section to 'What a Step Contains'
New Features
- Optimized plan generation that focuses on decisions rather than full code transcription
- Self-review mechanism to ensure plans remain lean and proportional to the spec
- Improved plan execution reliability with planted-defect probes
Full Release Notes
`writing-plans` produces leaner plans, faster. Plans now record the decisions an implementer needs (signatures, test assertions, the spec's values) instead of writing out the code. Some frontier models, including Opus 5.5, could get overzealous during plan writing and, with certain prompting, would sometimes try to implement the entire project while designing the plan. The new skill keeps planning focused on the plan. When we reproduced the original report, the scratch builds went away, and plans took a quarter of the time and about a third of the tokens. Thanks to Harper Reed for the report and session bundle. (#2333) ### Writing Plans - **A plan records decisions. It's not a transcript of the code.** "What a Step Contains" replaces the "No Placeholders" section. A test step names the test and its assertions. A code step gives the exact signature, the file, and the spec's values, and includes a body only for an algorithm those don't determine. A verification step gives the command and its passing output. A reference to another task goes through that task's Interfaces block. Placeholders are still called out as the opposite failure. (#2333) - **Self-review checks proportion.** The plan compares its own length to the spec's. A plan several times longer than the spec is a transcript, and when code blocks dominate, bodies get replaced with signatures and test assertions. (#2333) - **The plan's reader is described as capable:** an engineer who writes idiomatic code once they know the exact interface and test. This replaces "zero context, questionable taste." Steps are now sized as "one action with a checkable result" instead of "2-5 minutes." (#2333) - Every plan written by the new skill executed 9/9 against planted-defect probes on Sonnet 5, the same result as full-code plans. (#2333) - Removed `plan-document-reviewer-prompt.md`. Nothing referenced it. (#2333) ### Documentation - Removed `CLAUDE.md`. Claude Code now reads `AGENTS.md` directly, but only when no `CLAUDE.md` exists, so keeping the one-line pointer would have hidden the real guidelines.
Related Repositories
public-apis/public-apis★ 491,193
A collective list of free APIs
NousResearch/hermes-agent★ 256,048
The agent that grows with you
yt-dlp/yt-dlp★ 200,218
A feature-rich command-line audio/video downloader
microsoft/markitdown★ 191,539
Python tool for converting files and office documents to Markdown.
Significant-Gravitas/AutoGPT★ 187,701
AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters.