Skip to content

docs(landing): refresh for v1.15 + richer, less static page #96

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
jkyberneees merged 1 commit into from
Jul 22, 2026
Merged

docs(landing): refresh for v1.15 + richer, less static page #96

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
167 changes: 158 additions & 9 deletions docs/index.html
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -90,6 +90,57 @@
.hero .install-block .btn:hover { opacity: .85; }
.cta-block { margin-top: 0; }

/* ── Hero stats bar ── */
.hero .stats {
display: flex; justify-content: center; gap: 40px; margin-top: 44px;
flex-wrap: wrap;
}
.hero .stats .stat { text-align: center; }
.hero .stats .stat .v {
font-size: 1.5rem; font-weight: 800; color: var(--accent);
font-family: var(--font-mono); letter-spacing: -.02em;
}
.hero .stats .stat .k {
font-size: 0.75rem; color: var(--text-secondary); margin-top: 2px;
}
.hero .stats .stat a .v { transition: opacity .15s; }
.hero .stats .stat a:hover .v { opacity: .8; text-decoration: none; }

/* ── Section nav (sticky) ── */
.section-nav {
position: sticky; top: 0; z-index: 20;
display: flex; justify-content: center; gap: 4px; flex-wrap: wrap;
padding: 10px 16px;
background: color-mix(in srgb, var(--bg-primary) 88%, transparent);
backdrop-filter: blur(8px); -webkit-backdrop-filter: blur(8px);
border-bottom: 1px solid var(--border-subtle);
}
.section-nav a {
color: var(--text-secondary); font-size: 0.75rem; padding: 4px 10px;
border-radius: 6px; transition: all .15s;
}
.section-nav a:hover { color: var(--accent); background: var(--bg-card); text-decoration: none; }

/* ── "New" badge ── */
.badge-new {
display: inline-block; background: var(--green); color: #000;
font-size: 0.625rem; font-weight: 700; padding: 2px 7px; border-radius: 4px;
margin-left: 8px; vertical-align: middle; letter-spacing: .04em;
}

/* ── What's-new cards ── */
.whatsnew { display: grid; grid-template-columns: 1fr 1fr; gap: 20px; }
.whatsnew .card {
background: var(--bg-secondary); border: 1px solid var(--border-subtle); border-radius: 12px;
padding: 24px; border-left: 3px solid var(--accent);
}
.whatsnew .card h4 { font-size: 1rem; font-weight: 600; margin-bottom: 6px; }
.whatsnew .card h4 .ver { color: var(--text-secondary); font-weight: 400; font-size: 0.75rem; font-family: var(--font-mono); margin-left: 8px; }
.whatsnew .card p { font-size: 0.8125rem; line-height: 1.5; }
.whatsnew .card p code { font-size: 0.75rem; }
.whatsnew .card a { font-size: 0.8125rem; }
@media (max-width: 700px) { .whatsnew { grid-template-columns: 1fr; } }

/* ── Copy button ── */
.copy-btn {
background: transparent; border: 1px solid var(--border-subtle); color: var(--text-secondary);
Expand Down Expand Up @@ -420,7 +471,7 @@
"name": "Which LLM providers does odek support?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Any OpenAI-compatible endpoint: Deepseek, OpenAI, Anthropic, Ollama, vLLM, Groq, Together, and Fireworks. odek is provider-agnostic, so you are not locked into a single vendor."
"text": "Any OpenAI-compatible endpoint: DeepSeek, OpenAI, Anthropic, Ollama, vLLM, Groq, Together, and Fireworks. odek is provider-agnostic, so you are not locked into a single vendor."
}
},
{
Expand Down Expand Up @@ -468,6 +519,14 @@ <h1>odek</h1>
<a class="btn" href="#install">Get Started →</a>
</div>

<div class="stats">
<div class="stat"><div class="v">~12 MB</div><div class="k">static binary</div></div>
<div class="stat"><div class="v">80.3%</div><div class="k">AIEB v2.0 score</div></div>
<div class="stat"><div class="v">25+</div><div class="k">built-in tools</div></div>
<div class="stat"><div class="v">4</div><div class="k">interfaces: CLI · REPL · Web · Telegram</div></div>
<div class="stat"><a href="https://github.com/BackendStack21/odek/releases/tag/v1.15.3"><div class="v">v1.15.3</div><div class="k">latest release ↗</div></a></div>
</div>

<!-- ═══ Hero carousel — zero → Telegram bot ═══ -->
<div class="carousel hero-carousel">
<div class="bar">
Expand Down Expand Up @@ -659,7 +718,7 @@ <h1>odek</h1>
<span class="tg-sender">Principal</span>
<div class="tg-bbl">
<div class="tg-reply">Hi</div>
<div class="tg-code">✅ Done · 1 turn · 10095 in / 155 out · cache: 0cr+0rd+0ct · 2s — tools: none</div>
<div class="tg-code">✅ Done · 1 turn · 10095 in / 155 out · cache: 8143 write / 0 read / 0 total · 2s — tools: none</div>
</div>
<span class="tg-time">13:10</span>
</div>
Expand All @@ -677,6 +736,52 @@ <h1>odek</h1>

<div class="container">

<nav class="section-nav" aria-label="Section navigation">
<a href="#whats-new">What's new</a>
<a href="#five-whys">Why</a>
<a href="#demos">Demos</a>
<a href="#features">Features</a>
<a href="#install">Install</a>
<a href="#cheatsheet">Cheatsheet</a>
<a href="#docs">Docs</a>
</nav>

<!-- ═══ What's new in v1.15 ═══ -->
<section id="whats-new">
<h2>What's new in <span class="hl">v1.15</span></h2>
<p style="margin-bottom:20px;color:var(--text-secondary);font-size:0.9375rem">
The biggest release since Extended Memory: the agent now anticipates, and cleans up after itself.
Full notes in the <a href="https://github.com/BackendStack21/odek/releases/tag/v1.15.0">v1.15.0 release</a>.
</p>

<div class="whatsnew">
<div class="card">
<h4>💡 Proactive Engagement <span class="ver">v1.15.0</span></h4>
<p>The agent ends answers with <strong>follow-up suggestions</strong> it already predicted, tracks your <strong>open loops</strong> (unanswered questions, stated goals), and can send <strong>opt-in proactive nudges</strong> — with hard anti-annoyance caps so it never becomes noise. Scheduled jobs are memory-wired, so a cron task can analyze your memory and sessions. <a href="https://github.com/BackendStack21/odek/blob/main/docs/EXTENDED_MEMORY.md#p6--proactive-engagement">docs ↗</a></p>
</div>
<div class="card">
<h4>🧹 Storage Maintenance <span class="ver">v1.15.0</span></h4>
<p>A janitor goroutine keeps long-lived agents (Telegram bot, <code>serve</code>) clean forever: session/audit/log/plan/media retention driven by an operator-only <code>maintenance</code> config. <code>odek cleanup --dry-run</code> previews exactly what would be deleted before anything is touched. <a href="https://github.com/BackendStack21/odek/blob/main/docs/MAINTENANCE.md">docs ↗</a></p>
</div>
<div class="card">
<h4>🧠 Memory overhaul <span class="ver">v1.15.0</span></h4>
<p>Semantic dedup of stored atoms, relevance-preserving recall ranking, atom consolidation, index staleness detection, and <code>odek memory extended stats</code> for operators. Guard-rejected memories go to quarantine for human review instead of vanishing.</p>
</div>
<div class="card">
<h4>🛡️ PIGuard + guard fixes <span class="ver">v1.15.x</span></h4>
<p>The prompt-injection sidecar got its score semantics fixed (confident-benign memories no longer rejected), a JSON-faithful HTTP gateway, NDJSON socket transport, and Apple-Silicon-safe builds. Skill auto-learn is hardened against project-repo injection.</p>
</div>
<div class="card">
<h4>🗣️ Multilingual voice (STT) <span class="ver">v1.15.2</span></h4>
<p>Docker images ship the multilingual <code>ggml-small</code> whisper model — usable <strong>German, English, and Spanish</strong> voice transcription out of the box (was effectively English-only).</p>
</div>
<div class="card">
<h4>📊 Honest cache metrics <span class="ver">v1.15.3</span></h4>
<p>DeepSeek's native cache fields are parsed, and providers that report nothing show <code>cache: n/a</code> instead of misleading zeros.</p>
</div>
</div>
</section>

<!-- ═══ Five Whys ═══ -->
<section id="five-whys">
<h2>The <span class="hl">Five Whys</span></h2>
Expand All @@ -691,7 +796,7 @@ <h2>The <span class="hl">Five Whys</span></h2>
</li>
<li>
<span class="q">Why?</span>
<span class="a">Because locking into one LLM provider is a strategic mistake. odek is <strong>provider-agnostic</strong> — Deepseek, OpenAI, Anthropic, Ollama, vLLM, Groq, Together, Fireworks. Any OpenAI-compatible endpoint.</span>
<span class="a">Because locking into one LLM provider is a strategic mistake. odek is <strong>provider-agnostic</strong> — DeepSeek, OpenAI, Anthropic, Ollama, vLLM, Groq, Together, Fireworks. Any OpenAI-compatible endpoint.</span>
</li>
<li>
<span class="q">Why?</span>
Expand Down Expand Up @@ -817,6 +922,21 @@ <h2>See it in <span class="hl">action</span></h2>
<span class="info">🏆 80.3% — highest published agent score</span>
</div>

<!-- Slide 8 -->
<div class="slide">
<span class="info"># Proactive engagement — the agent anticipates</span><br>
<span class="prompt">$ odek run --session "I plan to migrate Postgres 14 → 17 next quarter"</span><br>
<span class="output">Good call to plan early. Watch out for: logical replication</span><br>
<span class="output">slots, removed GUCs, and driver upgrades…</span><br>
<span class="info">── You might also want to ──</span><br>
<span class="output">• What are the breaking changes between PG 14 and 17?</span><br>
<span class="output">• How do I test the migration with minimal downtime?</span><br>
<span class="info"># … days later, via Telegram — an opt-in nudge:</span><br>
<span class="output">💡 Your Postgres migration goal has been quiet for a week —</span><br>
<span class="output"> want a pg_upgrade dry-run checklist?</span><br>
<span class="info"># capped: max/day + per-kind cooldown. Never noisy.</span>
</div>

</div>
<div class="nav"></div>
</div>
Expand Down Expand Up @@ -848,8 +968,26 @@ <h4>Self-Learning Skills</h4>

<div class="card">
<div class="icon">💾</div>
<h4>Persistent Memory</h4>
<p>Three tiers: facts (markdown on disk), session buffer, and searchable episodes. Merge-on-write via go-vector cosine similarity saves ~80% LLM calls. go-vector RP is ephemeral &mdash; rebuilt from text on every write, no embedding state to persist.</p>
<h4>Extended Memory</h4>
<p>Semantic, atomic memory: facts, preferences, goals extracted per turn, deduplicated by meaning, ranked by relevance and recency. Quarantine for untrusted content, associations between atoms, and a stats surface for operators. Legacy three-tier facts/buffer/episodes included too.</p>
</div>

<div class="card">
<div class="icon">💡</div>
<h4>Proactive Engagement <span class="badge-new">NEW</span></h4>
<p>The agent anticipates: follow-up suggestions after answers, open-loop tracking of your questions and goals, and opt-in proactive nudges with anti-annoyance caps. Scheduled jobs are memory-wired, so cron tasks can analyze memory and sessions.</p>
</div>

<div class="card">
<div class="icon">🧹</div>
<h4>Storage Maintenance <span class="badge-new">NEW</span></h4>
<p>Long-lived agents stay clean forever: a janitor handles session, audit-log, log, plan, and media retention on an operator-only config. <code>odek cleanup --dry-run</code> shows exactly what would be removed, first.</p>
</div>

<div class="card">
<div class="icon">🛡️</div>
<h4>PIGuard Sidecar</h4>
<p>Optional ML prompt-injection guard (ONNX, local) as a semantic second opinion on top of the rule-based scan — guarding memory writes, skill bodies, system prompts, and MCP descriptions. Docker Compose sidecar included.</p>
</div>

<div class="card">
Expand Down Expand Up @@ -891,7 +1029,7 @@ <h4>Parallel Tool Execution</h4>
<div class="card">
<div class="icon">💬</div>
<h4>Interaction Modes</h4>
<p>Three modes: <strong>engaging</strong> (emoji-rich LLM narration), <strong>enhance</strong> (narration persists after response), and <strong>verbose</strong> (raw tool traces). Per-tool trace messages with LLM reasoning content. Works across CLI, Telegram, and Web UI.</p>
<p>Four modes: <strong>engaging</strong> (emoji-rich LLM narration), <strong>enhance</strong> (narration persists after response), <strong>verbose</strong> (raw tool traces), and <strong>off</strong> (clean, machine-friendly). Per-tool trace messages with LLM reasoning content. Works across CLI, Telegram, and Web UI.</p>
</div>

<div class="card">
Expand Down Expand Up @@ -1013,7 +1151,13 @@ <h3>Commands</h3>
<li><span class="cmd">odek mcp</span><span class="desc">Start MCP server (stdio)</span></li>
<li><span class="cmd">odek mcp --sse-addr :8081</span><span class="desc">Start MCP server (SSE)</span></li>
<li><span class="cmd">odek telegram</span><span class="desc">Start Telegram bot</span></li>
<li><span class="cmd">odek run --deliver <task></span><span class="desc">Run task, deliver to Telegram</span></li>
<li><span class="cmd">odek schedule add &lt;task&gt;</span><span class="desc">Schedule a recurring agent task (cron)</span></li>
<li><span class="cmd">odek cleanup --dry-run</span><span class="desc">Preview storage maintenance, delete nothing</span></li>
<li><span class="cmd">odek cleanup</span><span class="desc">Run one storage maintenance sweep</span></li>
<li><span class="cmd">odek memory extended stats</span><span class="desc">Memory store, index, quarantine, recall health</span></li>
<li><span class="cmd">odek memory extended nudges</span><span class="desc">Preview proactive nudges (no cap consumed)</span></li>
<li><span class="cmd">odek skill list</span><span class="desc">List learned/installed skills</span></li>
<li><span class="cmd">odek run --deliver &lt;task&gt;</span><span class="desc">Run task, deliver to Telegram</span></li>
<li><span class="cmd">odek init [--global]</span><span class="desc">Create config file</span></li>
<li><span class="cmd">odek version</span><span class="desc">Show version</span></li>
</ul>
Expand All @@ -1028,6 +1172,7 @@ <h3>Key Flags</h3>
<li><span class="cmd">--no-learn</span><span class="desc">Disable skill learning</span></li>
<li><span class="cmd">--max-iter &lt;n&gt;</span><span class="desc">Max think→act cycles (default 90)</span></li>
<li><span class="cmd">--ctx &lt;files&gt; / -c</span><span class="desc">Attach files as context blocks (comma-separated)</span></li>
<li><span class="cmd">--memory-extended-enabled</span><span class="desc">Enable Extended Memory (opt-in)</span></li>
<li><span class="cmd">--system &lt;prompt&gt;</span><span class="desc">Override system prompt</span></li>
<li><span class="cmd">--no-color</span><span class="desc">Disable colored output</span></li>
<li><span class="cmd">--deliver</span><span class="desc">Deliver result to Telegram default chat</span></li>
Expand All @@ -1045,10 +1190,12 @@ <h2>Everything else is <span class="hl">documented</span></h2>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/CHEATSHEET.md">Cheat Sheet<span>Quick reference: commands, config, memory, subagents, env vars</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/API.md">Go SDK Guide<span>Import, Agent, Tools, memory, multi-turn sessions, examples</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/CONFIG.md">Configuration<span>Config files, env vars, priority chain</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/PROVIDERS.md">Providers & Models<span>Deepseek, OpenAI, Anthropic, Ollama, vLLM, more</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/PROVIDERS.md">Providers & Models<span>DeepSeek, OpenAI, Anthropic, Ollama, vLLM, more</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/CACHING.md">Prompt Caching<span>Cache markers for lower latency and cost on repeats</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/MCP.md">MCP (Two-Way)<span>Serve tools + connect to external MCP servers</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/MEMORY.md">Memory System<span>Three-tier facts, buffer, episodes</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/MEMORY.md">Memory System<span>Legacy facts, buffer, episodes</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/EXTENDED_MEMORY.md">Extended Memory<span>Semantic atoms, recall, quarantine, proactive engagement</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/MAINTENANCE.md">Storage Maintenance<span>Janitor, retention config, odek cleanup</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/SESSIONS.md">Multi-Turn Sessions<span>Save, resume, list, trim, cleanup</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/SANDBOXING.md">Sandboxing<span>Docker isolation, config, security model</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/DOCKER_COMPOSE_USER_GUIDE.md">Docker Compose<span>Run odek in Docker with Compose, step by step</span></a>
Expand All @@ -1058,6 +1205,8 @@ <h2>Everything else is <span class="hl">documented</span></h2>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/WEBUI.md">Web UI<span>odek serve, WebSocket protocol, @ resources</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/SCHEDULES.md">Scheduled Tasks<span>Native cron — run agent tasks and deliver results</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/TELEGRAM.md">Telegram Bot<span>Run agent tasks from Telegram: voice, photos, sessions</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/TOOL_SELECTION.md">Tool Selection<span>Enable/disable tools per profile, MCP naming</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/DAILY-WORKER.md">Daily Worker<span>odek as your daily chief-of-staff assistant</span></a>
<a href="https://github.com/BackendStack21/odek/blob/main/docs/DEVELOPMENT.md">Development<span>Building, testing, contributing</span></a>
</div>
</section>
Expand Down
Loading