You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
<p>This file contains API Key access tokens, refresh tokens, and expiry timestamps. It is automatically managed — you should not edit it manually.</p>
149
+
<p>This file contains OAuth access tokens, refresh tokens, and expiry timestamps. It is automatically managed — you should not edit it manually.</p>
OpenAnalyst CLI is an <strong>independent, open-source AI coding agent</strong> that connects to every major LLM provider through a single, unified terminal interface. Built from the ground up in Rust, it features a full full-screen terminal UI, multi-provider streaming, multi-agent orchestration, 51+ slash commands, 19 built-in tools, and MCP protocol support.
84
+
OpenAnalyst CLI is an <strong>independent, open-source AI coding agent</strong> that connects to every major LLM provider through a single, unified terminal interface. It features a full full-screen terminal UI, multi-provider streaming, multi-agent orchestration, 51+ slash commands, 19 built-in tools, and MCP protocol support.
Copy file name to clipboardExpand all lines: Documentation/tui.html
+1-1Lines changed: 1 addition & 1 deletion
Original file line number
Diff line number
Diff line change
@@ -90,7 +90,7 @@ <h4>Status Bar</h4>
90
90
<divclass="icon">⌨</div>
91
91
<divclass="content">
92
92
<h4>Input Box</h4>
93
-
<p>Text input with vim-mode support (via edtui), slash command autocomplete, and input history navigation. Right-side badges show: <strong>permission mode</strong> (when not Default), <strong>active agent</strong> (purple badge), and <strong>git branch</strong> (blue badge). Border color changes with permission level.</p>
93
+
<p>Text input with vim-mode support (with built-in vim engine), slash command autocomplete, and input history navigation. Right-side badges show: <strong>permission mode</strong> (when not Default), <strong>active agent</strong> (purple badge), and <strong>git branch</strong> (blue badge). Border color changes with permission level.</p>
<divclass="section-sub">The most feature-rich AI agent CLI ever built. Native Rust binary, 7 LLM providers, 65+ commands, 24 tools, voice input, multi-agent orchestration, clipboard paste, /undo revert, and more.</div>
241
+
<divclass="section-sub">The most feature-rich AI agent CLI ever built. Native binary, 7 LLM providers, 65+ commands, 24 tools, voice input, multi-agent orchestration, clipboard paste, /undo revert, and more.</div>
243
242
244
243
<divclass="acc open" onclick="toggleAcc(this)">
245
244
<divclass="acc-header"><divclass="acc-icon" style="background:rgba(255,107,0,0.15);color:var(--accent)">⚡</div><divclass="acc-title">Smart Per-Action Model Routing <spanclass="tag tag-n">NEW</span></div><divclass="acc-arrow">▶</div></div>
246
245
<divclass="acc-body"><divclass="acc-content">Every prompt is auto-classified into <strong>explore</strong>, <strong>research</strong>, <strong>code</strong>, or <strong>write</strong> — each routed to the optimal model+effort. Explore tasks go to Haiku (1K thinking), coding goes to Opus (32K thinking). Save 80% on tokens without thinking about it.<br><br><codestyle="color:var(--cyan)">/effort code max</code> — per-category control<br><codestyle="color:var(--cyan)">/route</code> — view/edit the full routing table<br><codestyle="color:var(--cyan)">/effort high</code> — set all categories globally</div></div>
247
246
</div>
248
247
<divclass="acc" onclick="toggleAcc(this)">
249
248
<divclass="acc-header"><divclass="acc-icon" style="background:rgba(34,211,238,0.15);color:var(--cyan)">★</div><divclass="acc-title">Knowledge Base — Agentic RAG <spanclass="tag tag-n">NEW</span></div><divclass="acc-arrow">▶</div></div>
250
-
<divclass="acc-body"><divclass="acc-content"><codestyle="color:var(--cyan)">/knowledge best Meta Ads strategy for D2C brands</code><br><br>Powered by <strong>BGE-M3 1024-dim embeddings</strong> on A100 GPU, <strong>PostgreSQL pgvector</strong>, and <strong>Neo4j knowledge graph</strong>. The pipeline:<br><br>1. <strong>Local MOE intent classification</strong> — Rust-side, zero latency (strategic, procedural, factual, comparative, diagnostic, etc.)<br>2. <strong>API call</strong> to hosted AgenticRAG with intent hint<br>3. <strong>Hybrid search</strong> — pgvector cosine + PostgreSQL FTS + Neo4j graph expansion<br>4. <strong>RRF fusion</strong> — merges results from all sources<br>5. <strong>KnowledgeCard</strong> — tabbed, collapsible results with abstracted category labels<br>6. <strong>Feedback</strong> — inline thumbs-up/down + <code>/feedback</code> corrections<br>7. <strong>Local cache</strong> — instant replay from <code>.openanalyst/knowledge/</code><br><br><strong>No raw course names exposed</strong> — results show "Ads Strategy", "AI & Machine Learning", etc.<br>Set <code>OPENANALYST_API_KEY=oa_your_key</code> to access.</div></div>
249
+
<divclass="acc-body"><divclass="acc-content"><codestyle="color:var(--cyan)">/knowledge best Meta Ads strategy for D2C brands</code><br><br>Powered by <strong>BGE-M3 1024-dim embeddings</strong> on A100 GPU, <strong>PostgreSQL pgvector</strong>, and <strong>Neo4j knowledge graph</strong>. The pipeline:<br><br>1. <strong>Local MOE intent classification</strong> — local, zero latency (strategic, procedural, factual, comparative, diagnostic, etc.)<br>2. <strong>API call</strong> to hosted AgenticRAG with intent hint<br>3. <strong>Hybrid search</strong> — pgvector cosine + PostgreSQL FTS + Neo4j graph expansion<br>4. <strong>RRF fusion</strong> — merges results from all sources<br>5. <strong>KnowledgeCard</strong> — tabbed, collapsible results with abstracted category labels<br>6. <strong>Feedback</strong> — inline thumbs-up/down + <code>/feedback</code> corrections<br>7. <strong>Local cache</strong> — instant replay from <code>.openanalyst/knowledge/</code><br><br><strong>No raw course names exposed</strong> — results show "Ads Strategy", "AI & Machine Learning", etc.<br>Set <code>OPENANALYST_API_KEY=oa_your_key</code> to access.</div></div>
<divclass="acc-body"><divclass="acc-content">Inspired by Andrej Karpathy’s agent philosophy: simple loop, good model, basic tools, verifiable criteria. The autonomous agent runs a <strong>think→act→observe→verify</strong> loop without user interaction.<br><br><codestyle="color:var(--cyan)">/openanalyst fix all failing tests --criteria "cargo test"</code><br><codestyle="color:var(--cyan)">/oa refactor auth to async --goal "all async" --criteria "cargo build"</code><br><codestyle="color:var(--cyan)">/oa add caching --max-turns 20</code><br><br>Optional params: <code>--goal</code> (description), <code>--criteria</code> (shell command to verify), <code>--max-turns</code> (default 30). Criteria commands have 60s timeout. Turn budget enforced as hard limit.</div></div>
277
+
<divclass="acc-body"><divclass="acc-content">Inspired by Andrej Karpathy’s agent philosophy: simple loop, good model, basic tools, verifiable criteria. The autonomous agent runs a <strong>think→act→observe→verify</strong> loop without user interaction.<br><br><codestyle="color:var(--cyan)">/openanalyst fix all failing tests --criteria "npm test"</code><br><codestyle="color:var(--cyan)">/oa refactor auth to async --goal "all async" --criteria "npm run build"</code><br><codestyle="color:var(--cyan)">/oa add caching --max-turns 20</code><br><br>Optional params: <code>--goal</code> (description), <code>--criteria</code> (shell command to verify), <code>--max-turns</code> (default 30). Criteria commands have 60s timeout. Turn budget enforced as hard limit.</div></div>
<divclass="acc-body"><divclass="acc-content">Single Rust binary. No Node.js, no Python, no Docker. Native on macOS (Intel + Apple Silicon), Linux (x64 + ARM), and Windows. Fast startup, low memory. ~13MB release binary.</div></div>
297
+
<divclass="acc-body"><divclass="acc-content">Single native binary. No Node.js, no Python, no Docker. Native on macOS (Intel + Apple Silicon), Linux (x64 + ARM), and Windows. Fast startup, low memory. ~18MB binary.</div></div>
0 commit comments