We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
v0.13.0 refresh: MCP 13-tool surface, status consolidation, LONGHAND_DATA_DIR, real roadmap, current counts
Update for v0.6.0: reconcile command, staleness surface, upgrade path
Refresh Home with v0.5.10 stats and add Upgrading section - Bumps stale stats (v0.5.8 → v0.5.10, 140 → 170 tests) - Replaces the one-line "To upgrade later" note with a proper Upgrading section explaining the v0.5.9 → v0.5.10 reanalyze step so existing users can rebuild their episode store in place
Add Token budget to MCP reference + refresh Home stale stats Mirrors the main repo's new README "Token budget" section so the wiki stays aligned. Also fixes stale Home.md claims: v0.5.1 → v0.5.8, 86 tests → 140, 16 MCP tools → 17. Token budget in MCP-Tools-Reference sits above the decision tree because "does it cost a lot of tokens" is the first question agents and users hit. Includes per-tool caps table, the comparison (raw JSONL / bigger-window / summarizer-tools), and tuning knobs (max_chars, summary_only, filters). Co-Authored-By: Claude <noreply@anthropic.com>
docs: update install instructions to pip install longhand Longhand is now on PyPI (v0.5.3+), so the recommended install is 'pip install longhand' rather than git clone + editable mode. The from-source path is preserved as a secondary 'Developer install' section for contributors.
Initial wiki: Home, Getting Started, MCP Tools Reference + stubs Home, Getting-Started, and MCP-Tools-Reference are fully drafted. CLI-Reference, Architecture, Configuration, Troubleshooting, Security, and Roadmap are scaffolded with section headers and TODO markers. _Sidebar provides persistent left-nav grouped by use case. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Initial Home page