diff options
Diffstat (limited to 'help.html')
| -rw-r--r-- | help.html | 87 |
1 files changed, 79 insertions, 8 deletions
@@ -231,9 +231,9 @@ footer .tagline{color:var(--accent2);margin-top:10px;font-weight:500} <h1>WebX11 <span>Docs</span></h1> </div> <div class="header-meta"> - <span class="badge badge-v">v4.4</span> + <span class="badge badge-v">v4.5</span> <span class="badge badge-new">28 core · 17 ext</span> - <span style="color:var(--text2)">150+ commands</span> + <span style="color:var(--text2)">160+ commands</span> </div> </div> </header> @@ -490,6 +490,7 @@ diff a.txt b.txt</pre> <tr><td><code>clear</code> / <code>cls</code></td><td>Clear terminal</td><td><code>cls</code></td></tr> <tr><td><code>echo</code></td><td>Print text</td><td><code>echo Hello</code></td></tr> <tr><td><code>mem</code></td><td>Show memory usage</td><td><code>mem</code></td></tr> + <tr><td><code>config</code></td><td>View/set terminal config (max-lines, history, pipe-buffer)</td><td><code>config terminal-max-lines 500</code></td></tr> <tr><td><code>status</code></td><td>System status</td><td><code>status</code></td></tr> <tr><td><code>font-size</code></td><td>Change font size (8–32)</td><td><code>font-size 16</code></td></tr> <tr><td><code>load-local</code></td><td>Load module from VFS or URL</td><td><code>load-local modules/ext/calc.js</code></td></tr> @@ -521,6 +522,7 @@ diff a.txt b.txt</pre> <tr><td><code>ai-deny</code></td><td>Deny pending destructive action</td><td><code>ai-deny</code></td></tr> <tr><td><code>ai-stop</code></td><td>Cancel running AI request</td><td><code>ai-stop</code></td></tr> <tr><td><code>ai-compress</code></td><td>Manually compress conversation context</td><td><code>ai-compress</code></td></tr> + <tr><td><code>ai-config</code></td><td>View/set AI loop config (max-turns, max-context, max-context-tokens, confirm-timeout, chat-lock-timeout, skill-debounce, trivial-optimization)</td><td><code>ai-config max-turns 12</code></td></tr> <tr><td><code>debug</code></td><td>Toggle AI debug mode</td><td><code>debug</code></td></tr> </table></div> @@ -1018,6 +1020,43 @@ script greet.sh</pre> </table></div> </section> + <!-- ── Security ── --> + <section id="security"> + <h2>🔒 Security & Guardrails</h2> + <div class="callout callout-warning"> + <span class="callout-title">Protecting Your Terminal</span> + Three layered protections — guardrails block dangerous patterns, confidence escalation requires AI to self-assess, shadow VFS previews filesystem impact. + </div> + + <h3>Guardrail Policy</h3> + <p>Pre-execution guardrails intercept AI tool calls and block or warn on patterns matching dangerous operations:</p> + <div class="table-wrap"><table> + <tr><th>Rule</th><th>Description</th><th>Severity</th></tr> + <tr><td><code>rm-rf-block</code></td><td>Prevent recursive force deletes</td><td>🔴 deny</td></tr> + <tr><td><code>curl-pipe-bash</code></td><td>Block curl-to-bash execution</td><td>🔴 deny</td></tr> + <tr><td><code>rm-outer-vfs</code></td><td>Block deletion of sensitive metadata files</td><td>🔴 deny</td></tr> + <tr><td><code>mass-delete</code></td><td>Block mass-delete operations</td><td>🔴 deny</td></tr> + <tr><td><code>chmod-777</code></td><td>Block world-writable permissions</td><td>🔴 deny</td></tr> + <tr><td><code>eval-danger</code></td><td>Block dynamic code execution (eval/exec/system)</td><td>🔴 deny</td></tr> + <tr><td><code>external-dl</code></td><td>External file download warning</td><td>🟡 warn</td></tr> + <tr><td><code>env-overwrite</code></td><td>Modifying environment variables</td><td>🟡 warn</td></tr> + </table></div> + + <h3>Commands</h3> + <div class="table-wrap"><table> + <tr><th>Command</th><th>Description</th></tr> + <tr><td><code>guardrail</code></td><td>Show guardrail policy rules and violation count</td></tr> + <tr><td><code>guardrail-add <id> <severity> <pattern> <tools...></code></td><td>Add a custom guardrail rule (severity: deny/warn)</td></tr> + <tr><td><code>guardrail-remove <id></code></td><td>Remove a guardrail rule</td></tr> + <tr><td><code>shadow <command></code></td><td>Execute command in Shadow VFS — snapshot, run, diff changes</td></tr> + <tr><td><code>shadow-commit</code></td><td>Accept shadow VFS changes</td></tr> + <tr><td><code>shadow-rollback</code></td><td>Undo shadow VFS changes</td></tr> + </table></div> + + <h3>Confidence Escalation</h3> + <p>When the AI requests destructive actions, it must declare <code>Confidence: N%</code> in its reasoning. Below 90%, manual approval is required. If no confidence is declared, manual approval is required by default. This prevents the AI from making low-confidence destructive changes.</p> + </section> + <!-- ── Sessions ── --> <section id="sessions"> <h2>💬 Sessions</h2> @@ -1099,7 +1138,7 @@ vfs-import</pre> <tr><td><code>orchestrator-wake <id></code></td><td>Restore agent from VFS sleep state</td></tr> <tr><td><code>orchestrator-kill <id></code></td><td>Terminate an agent and revoke leases</td></tr> <tr><td><code>orchestrator-purge</code></td><td>Remove dead agents (>60s silent)</td></tr> - <tr><td><code>orchestrator-config <key> <value></code></td><td>Tune: spawn-limit, spawn-strategy, heartbeat-timeout, idle-timeout, restart-policy, a2a-polling</td></tr> + <tr><td><code>orchestrator-config <key> <value></code></td><td>Tune: spawn-limit, spawn-strategy, heartbeat-timeout, idle-timeout, restart-policy, a2a-polling, a2a-poll-interval, heartbeat-check, idle-sweep-interval, dead-purge-timeout (persisted)</td></tr> <tr><td><code>orchestrator-restart-policy <id> <policy></code></td><td>Set per-agent restart policy: on-failure, always, never</td></tr> <tr><td><code>lease-create <agent-id> [duration] [task]</code></td><td>Create a time-bound lease for a worker agent</td></tr> <tr><td><code>lease-list</code></td><td>Show active leases — agent tasks with status and TTL</td></tr> @@ -1135,6 +1174,7 @@ vfs-import</pre> <tr><td><code>nvfs-status</code></td><td>Show index stats and last scan time</td></tr> <tr><td><code>nvfs-resolve <concept></code></td><td>Find files related to a concept (exact + fuzzy)</td></tr> <tr><td><code>nvfs-tree [concept]</code></td><td>Show concept-to-file mapping tree</td></tr> + <tr><td><code>nvfs-config [key] [value]</code></td><td>View/set display limits (fuzzy-limit, tree-files, tree-concepts, resolve-files)</td></tr> </table></div> </section> @@ -1180,6 +1220,7 @@ vfs-import</pre> <tr><td><code>sandbox-status</code></td><td>Show worker pool health and task metrics</td></tr> <tr><td><code>sandbox-config <k> <v></code></td><td>Configure max-workers, min-workers</td></tr> <tr><td><code>sandbox-run <task></code></td><td>Execute a task in the sandbox pool</td></tr> + <tr><td><code>sandbox-kill <id></code></td><td>Terminate a running sandbox task</td></tr> </table></div> <p><code>orchestrate --sandboxed <task></code> routes delegation through the sandbox for full isolation.</p> </section> @@ -1194,9 +1235,10 @@ vfs-import</pre> <tr><td><code>memgraph-search <q></code></td><td>Find nodes by label, type, or id</td></tr> <tr><td><code>memgraph-stats</code></td><td>Most connected nodes and strongest edges</td></tr> <tr><td><code>memgraph-export</code></td><td>Export full graph as JSON to VFS</td></tr> - <tr><td><code>memgraph-prune</code></td><td>Remove stale low-weight edges (>30d, <3 uses)</td></tr> +<tr><td><code>memgraph-prune</code></td><td>Remove stale low-weight edges (>30d, <3 uses)</td></tr> + <tr><td><code>memgraph-learn <entity> <fact></code></td><td>Add a fact to the memory graph</td></tr> </table></div> - <p>Context is automatically injected into AI system prompts for cross-session awareness.</p> + <p>Context is automatically injected into AI prompts for cross-session awareness.</p> </section> <!-- ── World Model ── --> @@ -1237,6 +1279,15 @@ vfs-import</pre> <div class="table-wrap"><table> <tr><th>Command</th><th>Description</th></tr> <tr><td><code>security-level <0-3></code></td><td>Set active security level</td></tr> + <tr><td><code>security</code></td><td>Show full status: level, blocked count, destructive/system/whitelist sets, suspicious flags, registered commands</td></tr> + <tr><td><code>security-destructive list</code></td><td>Show destructive command set</td></tr> + <tr><td><code>security-destructive add|remove <cmd></code></td><td>Add/remove command from destructive set (persisted)</td></tr> + <tr><td><code>security-system list</code></td><td>Show system command set</td></tr> + <tr><td><code>security-system add|remove <cmd></code></td><td>Add/remove command from system set (persisted)</td></tr> + <tr><td><code>security-whitelist list</code></td><td>Show whitelist command set</td></tr> + <tr><td><code>security-whitelist add|remove <cmd></code></td><td>Add/remove command from whitelist (persisted)</td></tr> + <tr><td><code>security-flag list</code></td><td>Show suspicious flags</td></tr> + <tr><td><code>security-flag add|remove <flag></code></td><td>Add/remove suspicious flag (persisted, -- prefix optional)</td></tr> <tr><td><code>cmd-allow <command></code></td><td>Add command to allowlist (bypasses restrictions)</td></tr> <tr><td><code>cmd-deny <command></code></td><td>Add command to denylist (always blocked)</td></tr> </table></div> @@ -1275,7 +1326,7 @@ listmods</pre> <section id="ref"> <h2>⚡ Quick Reference</h2> <div class="quick-ref"> - <div class="quick-ref-title">WebX11 v4.3</div> + <div class="quick-ref-title">WebX11 v4.5</div> <div class="quick-ref-grid"> <div class="quick-ref-item"><span class="cmd">help</span><span class="desc">Show all commands</span></div> <div class="quick-ref-item"><span class="cmd">cls</span><span class="desc">Clear terminal</span></div> @@ -1327,9 +1378,29 @@ listmods</pre> <h2>📋 Changelog</h2> <div class="version-item"> - <div class="v-header"><span class="v-num">v4.3.1 — June 2026</span><span class="v-date">2026-06-03</span></div> + <div class="v-header"><span class="v-num">v4.5 — June 2026</span><span class="v-date">2026-06-06</span></div> + <ul> + <li><strong>Security Config CLI</strong> — <code>security-destructive</code>, <code>security-system</code>, <code>security-whitelist</code>, <code>security-flag</code> add|remove|list with localStorage persistence and boot loading</li> + <li><strong>Enhanced security command</strong> — Shows all sets (destructive, system, whitelist, flags) plus registered command count</li> + <li><strong>AI loop config</strong> — <code>ai-config max-turns|max-context|trivial-optimization</code> with persistence across sessions</li> + <li><strong>Unified config CLI</strong> — <code>config</code> command with 4 terminal tuning keys (max-lines, history, pipe-buffer), all persisted in localStorage</li> + <li><strong>Extended ai-config</strong> — New keys: <code>max-context-tokens</code>, <code>confirm-timeout</code>, <code>chat-lock-timeout</code>, <code>skill-debounce</code> with range validation</li> + <li><strong>OllamaModule consolidated</strong> — Removed duplicate <code>ollamamodels</code> code path (redirects to <code>models</code>), centralized endpoint via <code>ollamaManager.endpoint</code></li> + <li><strong>Neural VFS async indexing</strong> — Chunked processing with progress percentage, no UI freeze on large VFS. Improved <code>_summarize</code> (section heading → paragraph → sentence → fallback)</li> + <li><strong>Neural VFS display-limit config</strong> — <code>nvfs-config</code> command with 4 keys (fuzzy-limit, tree-files, tree-concepts, resolve-files), persisted</li> + <li><strong>Orchestrator event bus</strong> — In-memory pub/sub for instant A2A delivery (VFS polling remains as fallback). Improved <code>_deriveCapabilities</code> with 14 keyword mappings</li> + <li><strong>Orchestrator extended config</strong> — 4 new interval keys (a2a-poll-interval, heartbeat-check, idle-sweep-interval, dead-purge-timeout) + all config now persists across sessions</li> + <li>Cache: <code>2026-06-06-v4.5</code> — hard refresh required</li> + </ul> + </div> + + <div class="version-item"> + <div class="v-header"><span class="v-num">v4.4 — June 2026</span><span class="v-date">2026-06-05</span></div> <ul> - <li><strong>Neural VFS</strong> — Semantic concept indexing layer: <code>nvfs-index</code>, <code>nvfs-resolve</code>, <code>nvfs-status</code>, <code>nvfs-tree</code></li> + <li><strong>v4.4 complete</strong> — 46 modules, 24 skills, full A2A lifecycle, sandboxed execution, memory graph, world model, health monitoring</li> + <li><strong>A2A Mailbox Transport</strong> — <code>a2a-send</code>, <code>a2a-list</code>, 500ms polling with envelope routing (request/response/broadcast/heartbeat)</li> + <li><strong>Autonomous sub-agent spawning</strong> — auto-role derivation, sleep/reanimate VFS serialization</li> + <li><strong>Sandboxed execution</strong> — WebWorker pool <code>sandbox-run</code>, <code>sandbox-kill</code> with inline fallback for headless</li> <li><strong>A2A Protocol</strong> — Agent contract system: <code>contract-register</code>, <code>contract-list</code>, schema validation in orchestrator</li> <li><strong>Self-Optimization Telemetry</strong> — <code>system-health</code> command showing tool latency, VFS stats, error rates</li> <li><strong>AI tool: nvfs_resolve</strong> — AI can search VFS by concept</li> |