v6.3.0
microsoft/markitdownv6.3.0Aug 12, 2026by obra
Upgrading?
AI Summary
This release enhances harness support for Devin CLI and Hermes Agent, improves Subagent-Driven Development (SDD) conflict resolution and task batching, and fixes Windows compatibility issues with git worktrees and graph rendering.
Key Highlights
- Devin CLI plugin installation now works with auto-triggering skills.
- SDD controllers resolve plan conflicts without stalling and batch small same-shape tasks.
- Git worktree removal no longer destroys untracked files.
- SessionStart hook dispatches via Git Bash on Windows.
- Codex subagent waits are now event-driven for better performance.
New Features
- Hermes Agent skill registration and bootstrap loading.
- Batched same-shape task dispatch for subagents.
- SDD plans carry a Spec pointer for conflict resolution.
- Windows compatibility fixes for render-graphs.js and hook dispatching.
Full Release Notes
### Harness Support - **Devin CLI**: `devin plugins install obra/superpowers` now works, and skills auto-trigger at session start. (#1995) - **Hermes Agent**: install from a git clone; skills register with Hermes' native loader and the bootstrap loads on the first turn. (#1922, #2025) - **Grok Build CLI** added to the install docs. (#1919) ### Brainstorming - **Ceremony now scales to the task.** Requests are classified as spike, bounded, or architectural; small tasks skip the two-document ritual. Every path still stops for your approval before implementation. (#2063) ### Subagent-Driven Development - **Controllers no longer stall on plan conflicts.** Non-catastrophic conflicts and ambiguities get a recorded ruling and work continues; only destructive or irreversible actions still stop for a human. One donated session had sat blocked for almost nine hours on a question the controller could have decided. (#2077) - **The pre-dispatch conflict scan records its checks in the ledger** instead of just asserting the plan is clean. (#2080) - **Small same-shape tasks batch into one dispatch**, cutting subagent cost sharply on micro-task plans; batch reviews verify every file in the brief made it into the diff. (#2078) - **Implementers and reviewers may not spawn their own subagents**, which was producing duplicate reviews. (#2059) - **Plans carry a `Spec:` pointer** and SDD reads the spec at setup, so plan conflicts get resolved against the design instead of guessed at. (#2086) - Reviewers re-read evidence they find illegible instead of re-running the test suite (#2089), and circuit-breaker rulings now show up in the Finish report. ### Codex - Subagent waits are event-driven instead of poll-heavy, spawns pin model and reasoning effort explicitly, and the multi-agent reference is corrected against Codex source. (#2060, #2061, #2062) ### Finishing a Development Branch - **Worktree removal no longer destroys untracked files.** When `git worktree remove` refuses because the tree holds uncommitted work, the skill stops, names the files, and asks — instead of reaching for `--force`. (#2016, #1223, #2024) ### Fixes - `render-graphs.js` in writing-skills works on Windows. - Corrected Copilot CLI backgrounding guidance for Windows. (#1929, #2006) - `bump-version.sh` covers the Hermes manifest. ### Documentation - README: added a table of contents and reorganized Getting Started.
Related Repositories
A collective list of free APIs
The agent that grows with you
A feature-rich command-line audio/video downloader
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.
Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack.