content large_stringlengths 3 20.5k | url large_stringlengths 53 192 ⌀ | branch large_stringclasses 4
values | source large_stringclasses 51
values | embeddings listlengths 384 384 | score float64 -0.21 0.65 |
|---|---|---|---|---|---|
# Health Checks on macOS How to see whether the linked channel is healthy from the menu bar app. ## Menu bar - Status dot now reflects Baileys health: - Green: linked + socket opened recently. - Orange: connecting/retrying. - Red: logged out or probe failed. - Secondary line reads "linked · auth 12m" or shows the failu... | https://github.com/openclaw/openclaw/blob/main/docs/platforms/mac/health.md | main | opebclaw | [
-0.014649374410510063,
0.026660330593585968,
-0.038215622305870056,
0.012376360595226288,
0.027239523828029633,
-0.0785766988992691,
0.046028319746255875,
0.017238307744264603,
0.02175821177661419,
0.0508527010679245,
0.04551244527101517,
-0.008580435998737812,
-0.0037976952735334635,
0.06... | 0.129415 |
This flow lets the macOS app act as a full remote control for an OpenClaw gateway running on another host (desktop/server). The app can connect directly to trusted LAN/Tailnet gateway URLs or manage an SSH tunnel when the remote gateway is loopback-only. Health checks, Voice Wake forwarding, and Web Chat reuse the same... | https://github.com/openclaw/openclaw/blob/main/docs/platforms/mac/remote.md | main | opebclaw | [
-0.02395928092300892,
0.0013167459983378649,
-0.04864849895238876,
-0.006601135712116957,
-0.009686915203928947,
-0.0931345596909523,
-0.007464601192623377,
0.008937029168009758,
0.04384924843907356,
0.021758578717708588,
0.034308381378650665,
0.005028711166232824,
0.028805352747440338,
0.... | 0.092305 |
CLI isn't found remotely. 4. Health checks and Web Chat will now run through the selected transport automatically. ## Web Chat - \*\*SSH tunnel\*\*: Web Chat connects to the gateway over the forwarded WebSocket control port (default 18789). - \*\*Direct (ws/wss)\*\*: Web Chat connects straight to the configured gateway... | https://github.com/openclaw/openclaw/blob/main/docs/platforms/mac/remote.md | main | opebclaw | [
-0.012245704419910908,
-0.025367457419633865,
-0.03637530654668808,
-0.0716748908162117,
-0.003987966571003199,
-0.051215775310993195,
0.0667167454957962,
-0.01333974115550518,
-0.015389637090265751,
-0.01688273809850216,
0.03495606780052185,
-0.05122532695531845,
0.01374256331473589,
0.05... | 0.118642 |
Apply file changes using a structured patch format. This is ideal for multi-file or multi-hunk edits where a single `edit` call would be brittle. The tool accepts a single `input` string that wraps one or more file operations: ``` \*\*\* Begin Patch \*\*\* Add File: path/to/file.txt +line 1 +line 2 \*\*\* Update File: ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/apply-patch.md | main | opebclaw | [
-0.1128094494342804,
0.00963873416185379,
-0.001683706883341074,
0.005497076082974672,
-0.06659837812185287,
-0.008388804271817207,
-0.013857916928827763,
0.05382183939218521,
-0.05104728788137436,
0.10402201861143112,
0.024944430217146873,
0.03335769474506378,
0.004717536736279726,
-0.076... | 0.042007 |
The Gateway handles commands sent as standalone messages starting with `/`. Host-only bash commands use `! ` (with `/bash ` as an alias). When a conversation is bound to an ACP session, normal text routes to the ACP harness. Gateway management commands remain local: `/acp ...` always reaches the OpenClaw command handle... | https://github.com/openclaw/openclaw/blob/main/docs/tools/slash-commands.md | main | opebclaw | [
-0.01429735030978918,
-0.05645474046468735,
-0.004905578214675188,
0.07659757882356644,
-0.003227785462513566,
-0.08453052490949631,
0.01584666222333908,
0.023710662499070168,
0.0785992294549942,
0.041810259222984314,
0.038104698061943054,
-0.011421957984566689,
-0.008868137374520302,
-0.0... | 0.109112 |
the current session and start a fresh one | | `/reset [soft [message]]` | Reset the current session in place. `soft` keeps the transcript, drops reused CLI backend session ids, and reruns startup | | `/compact [instructions]` | Compact the session context. See [Compaction](/concepts/compaction) | | `/stop` | Abort the ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/slash-commands.md | main | opebclaw | [
-0.07396571338176727,
-0.0543966181576252,
-0.03775038197636604,
0.00823814794421196,
-0.020041421055793762,
-0.060884200036525726,
0.019257307052612305,
0.03448500856757164,
0.06413327157497406,
0.01079544983804226,
-0.013977834023535252,
0.0412178635597229,
0.0012755738571286201,
0.04457... | 0.08423 |
| `/usage off\|tokens\|full\|cost` | Control the per-response usage footer or print a local cost summary | | Command | Description | | --- | --- | | `/skill [input]` | Run a skill by name | | `/allowlist [list\|add\|remove] ...` | Manage allowlist entries. Text-only | | `/approve ` | Resolve exec or plugin approval pro... | https://github.com/openclaw/openclaw/blob/main/docs/tools/slash-commands.md | main | opebclaw | [
-0.06106593459844589,
0.025995925068855286,
-0.08129198849201202,
0.0713648647069931,
-0.004863106179982424,
0.001064905198290944,
0.09731034189462662,
0.10014168918132782,
-0.03568744659423828,
0.057446468621492386,
0.043819963932037354,
-0.09582985192537308,
0.019613180309534073,
0.01265... | 0.134985 |
when required args are omitted. Telegram and Slack show a button menu for commands with choices. Dynamic choices resolve against the target session model, so model- specific options like `/think` levels follow the session's `/model` override. ## `/tools` — what the agent can use now `/tools` answers a runtime question:... | https://github.com/openclaw/openclaw/blob/main/docs/tools/slash-commands.md | main | opebclaw | [
-0.049655355513095856,
-0.030905084684491158,
-0.019889095798134804,
0.09516652673482895,
-0.047575682401657104,
-0.017957933247089386,
0.011647106148302555,
0.07326312363147736,
-0.041539669036865234,
0.03184295818209648,
-0.00738535588607192,
-0.06518325209617615,
-0.02866262011229992,
0... | 0.102328 |
Slack reserves `/status`. Text `/status` still works in Slack messages. - Command-only messages from allowlisted senders are handled immediately (bypass queue + model). - Inline shortcuts (`/help`, `/commands`, `/status`, `/whoami`) also work embedded in normal messages and are stripped before the model sees the remain... | https://github.com/openclaw/openclaw/blob/main/docs/tools/slash-commands.md | main | opebclaw | [
-0.042384352535009384,
-0.007920023053884506,
0.009704803116619587,
0.04325208067893982,
-0.023155860602855682,
-0.02979201264679432,
0.06704061478376389,
0.024240562692284584,
0.003447099355980754,
0.05827029049396515,
0.06902985274791718,
-0.025649849325418472,
0.044740740209817886,
0.06... | 0.119034 |
Lobster is a workflow shell that lets OpenClaw run multi-step tool sequences as a single, deterministic operation with explicit approval checkpoints. Lobster is one authoring layer above detached background work. For flow orchestration above individual tasks, see [Task Flow](/automation/taskflow) (`openclaw tasks flow`... | https://github.com/openclaw/openclaw/blob/main/docs/tools/lobster.md | main | opebclaw | [
-0.07052383571863174,
0.015656258910894394,
-0.004937479272484779,
-0.03397626429796219,
-0.008283089846372604,
-0.12222649902105331,
-0.04216987267136574,
-0.022671015933156013,
0.067123644053936,
-0.044634077697992325,
-0.0668984055519104,
-0.08131560683250427,
0.026396624743938446,
-0.0... | 0.173581 |
"id": "main", "tools": { "alsoAllow": ["llm-task"] } } ] } } ``` ### Important limitation: embedded Lobster vs `openclaw.invoke` The bundled Lobster plugin runs workflows \*\*in-process\*\* inside the gateway. In that embedded mode, `openclaw.invoke` does \*\*not\*\* automatically inherit a gateway URL/auth context for... | https://github.com/openclaw/openclaw/blob/main/docs/tools/lobster.md | main | opebclaw | [
-0.074149951338768,
0.010347668081521988,
-0.014286389574408531,
0.0011405091499909759,
-0.01273985207080841,
-0.09790072590112686,
-0.05871463194489479,
-0.010526487603783607,
0.013174920342862606,
-0.06959701329469681,
-0.03133116662502289,
-0.0853857770562172,
0.07370539754629135,
0.023... | 0.048442 |
One workflow. Deterministic. Safe. ## Tool parameters ### `run` Run a pipeline in tool mode. ```json { "action": "run", "pipeline": "gog.gmail.search --query 'newer\_than:1d' | email.triage", "cwd": "workspace", "timeoutMs": 30000, "maxStdoutBytes": 512000 } ``` Run a workflow file with args: ```json { "action": "run",... | https://github.com/openclaw/openclaw/blob/main/docs/tools/lobster.md | main | opebclaw | [
-0.08412768691778183,
0.011924686841666698,
-0.00792800635099411,
0.0010940536158159375,
0.015966905280947685,
-0.06051254644989967,
-0.04407058283686638,
0.013499892316758633,
0.003386562690138817,
-0.025559455156326294,
0.02340681292116642,
-0.03263897821307182,
-0.0040718140080571175,
-... | 0.076656 |
Permission modes decide how much authority an agent has before it can run host commands, write files, or ask a backend harness for extra access. Start with `tools.exec.mode: "auto"` when you want OpenClaw to use allowlists first, then Codex native auto-review or a human approval route for misses. Permission mode is sep... | https://github.com/openclaw/openclaw/blob/main/docs/tools/permission-modes.md | main | opebclaw | [
0.02100948803126812,
0.008239264599978924,
-0.12322257459163666,
0.05603788048028946,
0.05487726628780365,
-0.10719602555036545,
0.0019096324685961008,
-0.04376067593693733,
-0.03761248290538788,
0.013035241514444351,
0.03619878739118576,
-0.023504452779889107,
-0.013920084573328495,
0.005... | 0.06274 |
auto-review before humans | `tools.exec.mode: "auto"` | | Skip host exec approvals entirely | `tools.exec.mode: "full"` plus matching host approvals file | | Make non-interactive ACPX sessions write/exec | `plugins.entries.acpx.config.permissionMode: "approve-all"` | If a command still prompts or fails after changing m... | https://github.com/openclaw/openclaw/blob/main/docs/tools/permission-modes.md | main | opebclaw | [
-0.03576049581170082,
0.02500317245721817,
-0.04167721047997475,
0.0480768084526062,
0.04695630073547363,
-0.08312133699655533,
0.013578691519796848,
-0.04827653616666794,
-0.040312014520168304,
0.017346471548080444,
0.08248918503522873,
-0.059057071805000305,
0.00404296163469553,
0.032844... | 0.045617 |
Exec approvals are the \*\*companion app / node host guardrail\*\* for letting a sandboxed agent run commands on a real host (`gateway` or `node`). A safety interlock: commands are allowed only when policy + allowlist + (optional) user approval all agree. Exec approvals stack \*\*on top of\*\* tool policy and elevated ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals.md | main | opebclaw | [
-0.015561635605990887,
0.018666917458176613,
-0.043626271188259125,
0.012804188765585423,
0.0837457925081253,
-0.05996806174516678,
0.029095912352204323,
-0.034336794167757034,
-0.0036119623109698296,
0.051809873431921005,
0.05854516848921776,
-0.024916235357522964,
0.0028698756359517574,
... | 0.181818 |
"autoAllowSkills": false }, "agents": { "main": { "security": "allowlist", "ask": "on-miss", "askFallback": "deny", "autoAllowSkills": true, "allowlist": [ { "id": "B0C8C0B3-2C2D-4F8A-9A3C-5A4B3C2D1E0F", "pattern": "~/Projects/\*\*/bin/rg", "source": "allow-always", "commandText": "rg -n TODO", "lastUsedAt": 1737150000... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals.md | main | opebclaw | [
-0.05874136462807655,
0.08236786723136902,
-0.07938258349895477,
-0.008469099178910255,
0.023995965719223022,
-0.012828394770622253,
0.07177962362766266,
0.010820531286299229,
-0.03916880488395691,
0.06874804943799973,
-0.02041364274919033,
-0.09132522344589233,
-0.012728330679237843,
0.05... | 0.102713 |
make gateway routing a free override from a sandboxed session. A per-call `host=node` request is allowed from `auto`; `host=gateway` is only allowed from `auto` when no sandbox runtime is active. For a stable non-auto default, set `tools.exec.host` or use `/exec host=...` explicitly. CLI-backed providers that expose th... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals.md | main | opebclaw | [
-0.009205217473208904,
0.027254974469542503,
-0.06486109644174576,
0.01835034228861332,
0.000557075603865087,
-0.10584015399217606,
-0.05426325276494026,
-0.040169864892959595,
-0.05299339070916176,
-0.006479041650891304,
-0.0002946456370409578,
0.013322217389941216,
-0.0172786433249712,
-... | 0.002753 |
other.py` is an allowlist miss. If a path-only entry for the same binary is also present, unmatched arguments can still fall back to that path-only entry. Omit the path-only entry when the goal is to restrict the binary to the declared arguments. Entries saved by approval flows can use an internal separator format for ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals.md | main | opebclaw | [
-0.05235034227371216,
0.024794114753603935,
-0.028962302953004837,
0.03188140317797661,
0.0390513613820076,
-0.10778368264436722,
0.0003030858642887324,
-0.008802072145044804,
0.0036177949514240026,
-0.03665241599082947,
-0.003466404974460602,
-0.02216547168791294,
0.0014621770242229104,
0... | 0.012623 |
- `Exec running` (only if the command exceeds the running notice threshold). - `Exec finished`. These are posted to the agent's session after the node reports the event. Denied exec approvals are terminal for the host command itself: the command does not run. For main-agent async approvals with an originating session, ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals.md | main | opebclaw | [
-0.01088484562933445,
0.025147585198283195,
-0.055616650730371475,
0.08333773165941238,
0.06550886482000351,
-0.11402171105146408,
-0.04533613845705986,
-0.0632128119468689,
0.07293255627155304,
-0.014411445707082748,
0.02772187441587448,
-0.005576878320425749,
0.015647711232304573,
-0.022... | 0.063362 |
Plugins extend OpenClaw with channels, model providers, agent harnesses, tools, skills, speech, realtime transcription, voice, media understanding, generation, web fetch, web search, and other runtime capabilities. Use this page when you want to install a plugin, restart the Gateway, verify that the runtime loaded it, ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/plugin.md | main | opebclaw | [
-0.016537858173251152,
-0.13424521684646606,
-0.06347580254077911,
0.020127423107624054,
-0.04773775860667229,
-0.045871198177337646,
-0.0830351710319519,
-0.005601801443845034,
-0.05549454316496849,
-0.009248609654605389,
0.031062772497534752,
-0.027095945551991463,
-0.038267750293016434,
... | 0.066381 |
specs have special compatibility behavior. If the bare name matches a bundled plugin id, OpenClaw uses that bundled source. If it matches an official external plugin id, OpenClaw uses the official package catalog. Other ordinary bare package specs install through npm during the launch cutover. Raw `@openclaw/\*` packag... | https://github.com/openclaw/openclaw/blob/main/docs/tools/plugin.md | main | opebclaw | [
-0.08440456539392471,
-0.005812097806483507,
0.029474865645170212,
0.0486411526799202,
0.12210649251937866,
-0.06833506375551224,
-0.06252080947160721,
0.02265462838113308,
-0.0415513813495636,
-0.013281221501529217,
0.05369005724787712,
-0.046932246536016464,
-0.01868761144578457,
0.06440... | 0.004549 |
bundle compatibility boundary and [Building plugins](/plugins/building-plugins) for native plugin authoring. ## Plugin hooks Plugins can register hooks at runtime, but there are two different APIs with different jobs. - Use typed hooks via `api.on(...)` for runtime lifecycle hooks. This is the preferred surface for mid... | https://github.com/openclaw/openclaw/blob/main/docs/tools/plugin.md | main | opebclaw | [
-0.14263609051704407,
-0.07021523267030716,
-0.03262583911418915,
0.05603133514523506,
0.00200649444013834,
-0.042240556329488754,
-0.02494235336780548,
0.05023486912250519,
-0.07128164917230606,
0.004685830324888229,
-0.0002857510989997536,
-0.03149101510643959,
-0.04377729818224907,
0.05... | 0.149788 |
keys without stale-plugin evidence still fail validation so typos stay visible. For intentional channel replacement, the preferred plugin should declare `channelConfigs..preferOver` with the legacy or lower-priority plugin id. If both plugins are explicitly enabled, OpenClaw keeps that request and reports duplicate cha... | https://github.com/openclaw/openclaw/blob/main/docs/tools/plugin.md | main | opebclaw | [
-0.1158110722899437,
-0.030696634203195572,
0.015013436786830425,
0.01922234334051609,
0.05049113929271698,
-0.11371315270662308,
-0.06418009102344513,
-0.07178764790296555,
0.007295167073607445,
-0.004286099690943956,
0.05563557520508766,
0.030048182234168053,
-0.0011157068656757474,
0.01... | -0.014872 |
package metadata | https://github.com/openclaw/openclaw/blob/main/docs/tools/plugin.md | main | opebclaw | [
0.004901490639895201,
0.03918005898594856,
-0.06894972175359726,
-0.08067531138658524,
0.05188445374369621,
-0.017902536317706108,
0.06503480672836304,
0.023764632642269135,
-0.04639710113406181,
0.013874327763915062,
0.043327447026968,
-0.024586820974946022,
0.03835999220609665,
-0.030412... | 0.090735 |
Advanced exec-approval topics: the `safeBins` fast-path, interpreter/runtime binding, and approval-forwarding to chat channels (including native delivery). For the core policy and approval flow, see [Exec approvals](/tools/exec-approvals). ## Safe bins (stdin-only) `tools.exec.safeBins` defines a small list of \*\*stdi... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals-advanced.md | main | opebclaw | [
-0.06542856246232986,
-0.028553202748298645,
-0.06270212680101395,
-0.02825651317834854,
0.04449818283319473,
-0.0752946212887764,
0.09361681342124939,
0.019601378589868546,
-0.027707068249583244,
0.03434499353170395,
-0.011798116378486156,
-0.025839749723672867,
0.008069495670497417,
0.03... | 0.159368 |
requires an explicit approval. In strict mode, `allow-always` can still persist benign interpreter/script invocations, but inline-eval carriers are not persisted automatically. ### Safe bins versus allowlist | Topic | `tools.exec.safeBins` | Allowlist (`exec-approvals.json`) | | ---------------- | ---------------------... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals-advanced.md | main | opebclaw | [
-0.058808423578739166,
0.0007546354900114238,
-0.0592544749379158,
-0.005016612354665995,
0.059507858008146286,
-0.025098638609051704,
0.010335412807762623,
-0.052202530205249786,
-0.01963816024363041,
-0.0014744529034942389,
0.04427814483642578,
-0.02543872781097889,
-0.02145354077219963,
... | 0.08878 |
Denials without a resumable main session are either suppressed or reported through a safe direct route when one exists. - If a valid external delivery target exists (deliverable channel plus target `to`), followup delivery uses that channel. - In webchat-only or internal-session flows with no external target, followup ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals-advanced.md | main | opebclaw | [
-0.06299147009849548,
-0.0182273518294096,
0.07140853255987167,
0.004880608059465885,
0.044542256742715836,
-0.07736214995384216,
-0.0029623250011354685,
-0.0418279692530632,
0.0018560040043666959,
0.07827924191951752,
0.05071566253900528,
0.017277348786592484,
-0.05459074303507805,
0.0720... | 0.074492 |
agent-facing path. The agent should not also echo a duplicate plain chat `/approve` command unless the tool result says chat approvals are unavailable or manual approval is the only remaining path. If a native approval client is configured but no native runtime is active for the originating channel, OpenClaw keeps the ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals-advanced.md | main | opebclaw | [
-0.025591745972633362,
-0.07434424012899399,
-0.013286987319588661,
0.03148721158504486,
0.03936302289366722,
-0.09903494268655777,
-0.00005643080294248648,
-0.027998190373182297,
0.052225470542907715,
0.01313104946166277,
0.03425395488739014,
-0.038145240396261215,
-0.036043424159288406,
... | 0.103846 |
prompts include `com.openclaw.approval` custom event content on the first prompt event so OpenClaw-aware Matrix clients can read structured approval state while stock clients keep the plain-text `/approve` fallback - the requester does not need to be an approver - the originating chat can approve directly with `/approv... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals-advanced.md | main | opebclaw | [
-0.04651740565896034,
-0.043697185814380646,
-0.008095456287264824,
0.04229786619544029,
0.05251484364271164,
-0.1138978973031044,
0.0020028946455568075,
-0.03408808261156082,
0.04291070997714996,
0.022627152502536774,
0.03991173952817917,
-0.045760445296764374,
-0.05360456183552742,
0.047... | 0.106784 |
commands in that channel session. If the channel exposes explicit approval approvers, those approvers can authorize the `/approve` action even when they are not otherwise command-authorized in that session. Some channels are stricter. Discord, Telegram, Matrix, Slack native approval DMs, and similar native approval cli... | https://github.com/openclaw/openclaw/blob/main/docs/tools/exec-approvals-advanced.md | main | opebclaw | [
-0.03423680365085602,
-0.027328765019774437,
0.01824289746582508,
-0.022466400638222694,
0.0031356201507151127,
-0.05942812189459801,
0.08166419714689255,
-0.008597098290920258,
0.0033629490062594414,
0.06868825107812881,
0.002350613009184599,
-0.006930135656148195,
-0.00025802248273976147,
... | 0.105803 |
## What it does - Inline directive in any inbound body: `/t `, `/think:`, or `/thinking `. - Levels (aliases): `off | minimal | low | medium | high | xhigh | adaptive | max` - minimal → "think" - low → "think hard" - medium → "think harder" - high → "ultrathink" (max budget) - xhigh → "ultrathink+" (GPT-5.2+ and Codex ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/thinking.md | main | opebclaw | [
0.01758391037583351,
-0.04669586196541786,
-0.008814520202577114,
0.002234803279861808,
-0.02927270345389843,
-0.06133314222097397,
0.042109809815883636,
0.10034409910440445,
-0.052689727395772934,
0.023109285160899162,
-0.007714988198131323,
0.06060691550374031,
-0.0040820445865392685,
-0... | 0.089713 |
only supports binary thinking (`on`/`off`). Any non-`off` level is treated as `on` (mapped to `low`). - Moonshot (`moonshot/\*`) maps `/think off` to `thinking: { type: "disabled" }` and any non-`off` level to `thinking: { type: "enabled" }`. When thinking is enabled, Moonshot only accepts `tool\_choice` `auto|none`; O... | https://github.com/openclaw/openclaw/blob/main/docs/tools/thinking.md | main | opebclaw | [
0.05466651916503906,
-0.07436168193817139,
-0.03734324499964714,
0.02152862772345543,
0.06844434142112732,
-0.05550721287727356,
0.011226950213313103,
0.028417205438017845,
-0.02707093022763729,
0.020999109372496605,
0.06322455406188965,
-0.020952455699443817,
0.03901218995451927,
0.031678... | 0.044077 |
gateway/webchat clients need `operator.admin` to persist it. - Inline directive affects only that message; session/global defaults apply otherwise. - Send `/verbose` (or `/verbose:`) with no argument to see the current verbose level. - When verbose is on, agents that emit structured tool results send each tool call bac... | https://github.com/openclaw/openclaw/blob/main/docs/tools/thinking.md | main | opebclaw | [
0.0054184552282094955,
-0.033464692533016205,
-0.006024982314556837,
0.08675750344991684,
-0.02143084444105625,
-0.038003455847501755,
-0.012847057543694973,
0.008996561169624329,
-0.0002967326436191797,
0.017128759995102882,
0.025713158771395683,
-0.013341757468879223,
0.012997066602110863,... | 0.026692 |
the session override immediately via `sessions.patch`; it does not wait for the next send and it is not a one-shot `thinkingOnce` override. - The first option is always the clear-override choice. It shows `Inherited: `, including `Inherited: Off` when inherited thinking is disabled. - Explicit picker choices use their ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/thinking.md | main | opebclaw | [
-0.1342090368270874,
-0.05744215101003647,
0.05200474336743355,
0.03952536731958389,
-0.006101448554545641,
0.015360496938228607,
0.05212851241230965,
0.02579202875494957,
0.03272173926234245,
0.015290680341422558,
0.014837326481938362,
0.035113073885440826,
-0.08272263407707214,
0.0556956... | 0.002202 |
`tokenjuice` is an optional external plugin that compacts noisy `exec` and `bash` tool results after the command has already run. It changes the returned `tool\_result`, not the command itself. Tokenjuice does not rewrite shell input, rerun commands, or change exit codes. Today this applies to OpenClaw embedded runs an... | https://github.com/openclaw/openclaw/blob/main/docs/tools/tokenjuice.md | main | opebclaw | [
-0.052871767431497574,
0.04714786261320114,
-0.040267836302518845,
0.06712974607944489,
-0.054126013070344925,
-0.08149290829896927,
-0.0831596851348877,
-0.004392225295305252,
-0.013653702102601528,
-0.03537769615650177,
0.034118857234716415,
-0.008725832216441631,
0.02397290989756584,
-0... | 0.117945 |
OpenClaw supports MiniMax as a `web\_search` provider through the MiniMax Token Plan search API. It returns structured search results with titles, URLs, snippets, and related queries. ## Get a Token Plan credential Create or copy a MiniMax Token Plan key from [MiniMax Platform](https://platform.minimax.io/user-center/b... | https://github.com/openclaw/openclaw/blob/main/docs/tools/minimax-search.md | main | opebclaw | [
-0.0074555897153913975,
0.03877504914999008,
-0.0476190485060215,
0.03734955936670303,
-0.021464964374899864,
-0.042255669832229614,
-0.13417868316173553,
0.05854476988315582,
0.0010683223372325301,
-0.044103946536779404,
-0.011812497861683369,
0.011937055736780167,
0.060731712728738785,
-... | -0.023909 |
`code\_execution` runs sandboxed remote Python analysis on xAI's Responses API. It is registered by the bundled `xai` plugin (under the `tools` contract) and dispatches to the same `https://api.x.ai/v1/responses` endpoint used by `x\_search`. | Property | Value | | ------------------ | ---------------------------------... | https://github.com/openclaw/openclaw/blob/main/docs/tools/code-execution.md | main | opebclaw | [
-0.012360609136521816,
-0.0039305356331169605,
-0.08918863534927368,
0.04837403818964958,
0.03840036317706108,
-0.0791764184832573,
-0.01000964269042015,
-0.015853531658649445,
0.003433982143178582,
-0.02225077524781227,
0.041750699281692505,
-0.05586010217666626,
0.04762803390622139,
0.01... | 0.087671 |
analysis, not a persistent notebook session. - Do not assume access to local files or your workspace. - For fresh X data, use [`x\_search`](/tools/web#x\_search) first and pipe the result into `code\_execution`. ## Related Local shell execution on your machine or paired node. Allow/deny policy for shell execution. `web... | https://github.com/openclaw/openclaw/blob/main/docs/tools/code-execution.md | main | opebclaw | [
-0.03753026947379112,
0.07016110420227051,
-0.09094121307134628,
-0.007704196032136679,
0.02284640446305275,
-0.05020935460925102,
0.014557942748069763,
0.019005198031663895,
0.007965902797877789,
-0.028366856276988983,
0.02253367193043232,
0.022447578608989716,
0.01770688220858574,
-0.050... | 0.108057 |
`llm-task` is an \*\*optional plugin tool\*\* that runs a JSON-only LLM task and returns structured output (optionally validated against JSON Schema). This is ideal for workflow engines like Lobster: you can add a single LLM step without writing custom OpenClaw code for each workflow. ## Enable the plugin 1. Enable the... | https://github.com/openclaw/openclaw/blob/main/docs/tools/llm-task.md | main | opebclaw | [
-0.05025753751397133,
0.018079157918691635,
-0.023900358006358147,
0.014107812196016312,
0.03499052673578262,
-0.030799102038145065,
-0.08008228242397308,
-0.01660945825278759,
-0.004941950086504221,
-0.024264950305223465,
-0.02066483162343502,
-0.04657194763422012,
0.002148374915122986,
0... | 0.028524 |
In the common split-host setup, OpenClaw Gateway runs inside WSL2, Chrome runs on Windows, and browser control must cross the WSL2 and Windows boundary. The layered failure pattern from [issue #39369](https://github.com/openclaw/openclaw/issues/39369) means several independent problems can show up at once, which makes ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser-wsl2-windows-remote-cdp-troubleshooting.md | main | opebclaw | [
-0.002743671415373683,
-0.013818175531923771,
0.02673434466123581,
0.053129617124795914,
0.017110593616962433,
-0.08836336433887482,
-0.10926999896764755,
0.049784861505031586,
-0.015364128164947033,
-0.0933777317404747,
0.02703140489757061,
-0.07939507812261581,
-0.025390826165676117,
0.0... | -0.116174 |
profiles: { remote: { cdpUrl: "http://WINDOWS\_HOST\_OR\_IP:9222", attachOnly: true, color: "#00AA00", }, }, }, } ``` Notes: - use the WSL2-reachable address, not whatever only works on Windows - keep `attachOnly: true` for externally managed browsers - `cdpUrl` can be `http://`, `https://`, `ws://`, or `wss://` - use ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser-wsl2-windows-remote-cdp-troubleshooting.md | main | opebclaw | [
0.019057538360357285,
0.01477233786135912,
-0.055110733956098557,
0.09418235719203949,
0.03779854625463486,
-0.10039415210485458,
-0.08350373059511185,
0.02622002735733986,
-0.0017934147035703063,
-0.04824737086892128,
0.04935462400317192,
-0.05529019236564636,
-0.010256814770400524,
0.073... | -0.075631 |
`/btw` lets you ask a quick side question about the \*\*current session\*\* without turning that question into normal conversation history. `/side` is an alias. It is modeled after Claude Code's `/btw` behavior, but adapted to OpenClaw's Gateway and multi-channel architecture. ## What it does When you send: ```text /bt... | https://github.com/openclaw/openclaw/blob/main/docs/tools/btw.md | main | opebclaw | [
-0.09265587478876114,
0.02815929427742958,
-0.01430043950676918,
0.03900012746453285,
0.0011140448041260242,
-0.05792587250471115,
0.040381476283073425,
-0.01553428266197443,
0.010358992964029312,
-0.0028317745309323072,
-0.04144362360239029,
-0.04704966023564339,
0.03956031799316406,
-0.0... | 0.129134 |
behavior, but not yet a complete browser UX. ## When to use BTW Use `/btw` when you want: - a quick clarification about the current work, - a factual side answer while a long run is still in progress, - a temporary answer that should not become part of future session context. Examples: ```text /btw what file are we edi... | https://github.com/openclaw/openclaw/blob/main/docs/tools/btw.md | main | opebclaw | [
-0.0557088628411293,
0.007584839593619108,
0.04195064306259155,
0.04732092097401619,
0.07814311981201172,
-0.04292839393019676,
0.03560783341526985,
0.03574085980653763,
0.012681080028414726,
0.05467567965388298,
-0.01670370064675808,
-0.0352766327559948,
-0.05265085771679878,
0.0666542872... | 0.099657 |
OpenClaw can run a \*\*dedicated Chrome/Brave/Edge/Chromium profile\*\* that the agent controls. It is isolated from your personal browser and is managed through a small local control service inside the Gateway (loopback only). Beginner view: - Think of it as a \*\*separate, agent-only browser\*\*. - The `openclaw` pro... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser.md | main | opebclaw | [
-0.06115981563925743,
-0.021345308050513268,
-0.03893498331308365,
0.05463330075144768,
0.06414397805929184,
-0.049907900393009186,
-0.029686536639928818,
0.0692717432975769,
-0.057456761598587036,
-0.09983415901660919,
0.009046253748238087,
-0.03590375930070877,
0.04712173342704773,
-0.03... | 0.04101 |
{ plugins: { allow: ["telegram", "browser"], }, } ``` An explicit root `browser` block, for example `browser.enabled=true` or `browser.profiles.`, activates the bundled browser plugin even under a restrictive `plugins.allow`, matching channel config behavior. `plugins.entries.browser.enabled=true` and `tools.alsoAllow:... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser.md | main | opebclaw | [
-0.05189395695924759,
-0.02831963077187538,
-0.025692235678434372,
0.03374161943793297,
0.04801337420940399,
-0.06058124080300331,
-0.02848709002137184,
-0.01838899776339531,
-0.12345948815345764,
-0.05411602184176445,
0.0636162981390953,
0.007349969353526831,
0.01584758795797825,
0.093297... | 0.020358 |
for model fallbacks, timeouts, byte limits, profiles, and provider request settings. If the active main model already supports vision and no explicit image understanding model is configured, OpenClaw keeps the normal image result so the main model can read the screenshot directly. - Control service binds to loopback on... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser.md | main | opebclaw | [
0.0016426367219537497,
-0.03634801134467125,
-0.1005723848938942,
0.05470338091254234,
0.011068880558013916,
-0.054650988429784775,
-0.12347480654716492,
0.0397135354578495,
-0.008648434653878212,
-0.002993277506902814,
0.051352016627788544,
0.016734518110752106,
-0.03249165788292885,
-0.0... | 0.083633 |
or `WAYLAND\_DISPLAY`, local managed profiles default to headless automatically when neither the environment nor profile/global config explicitly chooses headed mode. `openclaw browser status --json` reports `headlessSource` as `env`, `profile`, `config`, `request`, `linux-display-fallback`, or `default`. - `OPENCLAW\_... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser.md | main | opebclaw | [
0.04908323660492897,
0.020907435566186905,
-0.05724151059985161,
0.031994909048080444,
0.061496127396821976,
-0.060554251074790955,
-0.07112332433462143,
0.08561784774065018,
-0.10140632092952728,
-0.058244768530130386,
0.07140874862670898,
-0.0457138866186142,
0.05285325646400452,
0.05594... | -0.00474 |
the auth when calling `/json/\*` endpoints and when connecting to the CDP WebSocket. Prefer environment variables or secrets managers for tokens instead of committing them to config files. ## Node browser proxy (zero-config default) If you run a \*\*node host\*\* on the machine that has your browser, OpenClaw can auto-... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser.md | main | opebclaw | [
-0.02282686159014702,
0.008589843288064003,
-0.06867468357086182,
0.03017275407910347,
0.00969597790390253,
-0.07060420513153076,
-0.0487915575504303,
0.02770528756082058,
-0.021456530317664146,
-0.03756085783243179,
-0.0010821996256709099,
0.007478801999241114,
0.00736733665689826,
0.0083... | 0.009295 |
falls back to a direct WebSocket handshake at the bare root. If the advertised WebSocket endpoint rejects the CDP handshake but the configured bare root accepts it, OpenClaw falls back to that root as well. This lets a bare `ws://` pointed at a local Chrome still connect, since Chrome only accepts WebSocket upgrades on... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser.md | main | opebclaw | [
-0.03094320371747017,
-0.01693817973136902,
0.008451416157186031,
0.0026356896851211786,
0.011699022725224495,
-0.10904531925916672,
-0.09116237610578537,
0.026451610028743744,
0.02460920251905918,
-0.03362104669213295,
0.010571228340268135,
0.02949477732181549,
-0.05095981806516647,
0.076... | 0.054735 |
\*\*remote\*\*: an explicit CDP URL (Chromium-based browser running elsewhere) - \*\*existing session\*\*: your existing Chrome profile via Chrome DevTools MCP auto-connect Defaults: - The `openclaw` profile is auto-created if missing. - The `user` profile is built-in for Chrome MCP existing-session attach. - Existing-... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser.md | main | opebclaw | [
-0.03444203734397888,
-0.03806447982788086,
0.023185888305306435,
-0.0017802466172724962,
0.017147734761238098,
-0.01517780777066946,
-0.009729665704071522,
0.048796795308589935,
-0.05692314729094505,
-0.06937160342931747,
0.019988978281617165,
-0.060606230050325394,
0.036037854850292206,
... | -0.004267 |
profile when the default `npx chrome-devtools-mcp@latest` flow is not what you want (offline hosts, pinned versions, vendored binaries): | Field | What it does | | ------------ | -------------------------------------------------------------------------------------------------------------------------- | | `mcpCommand` |... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser.md | main | opebclaw | [
-0.006589656230062246,
-0.04723136126995087,
0.007586795836687088,
0.010251878760755062,
0.06270530074834824,
-0.07335998117923737,
-0.039875634014606476,
0.04970403388142586,
-0.06434793770313263,
-0.032066460698843,
-0.009087459184229374,
-0.08930924534797668,
-0.041109904646873474,
0.00... | -0.01301 |
code paths. - \*\*CDP startup or readiness failure\*\* means OpenClaw cannot confirm that the browser control plane is healthy. - \*\*Navigation SSRF block\*\* means the browser control plane is healthy, but a page navigation target is rejected by policy. Common examples: - CDP startup or readiness failure: - `Chrome C... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser.md | main | opebclaw | [
0.023940201848745346,
-0.06042318418622017,
-0.09438273310661316,
-0.0008575012325309217,
0.04360935837030411,
-0.02288900502026081,
-0.033237144351005554,
0.04213956743478775,
-0.05350026115775108,
-0.015053628012537956,
0.08198828250169754,
0.02957485243678093,
-0.0039790901355445385,
0.... | 0.034766 |
[Agent Client Protocol (ACP)](https://agentclientprotocol.com/) sessions let OpenClaw run external coding harnesses (for example Claude Code, Cursor, Copilot, Droid, OpenClaw ACP, OpenCode, Gemini CLI, and other supported ACPX harnesses) through an ACP backend plugin. Each ACP session spawn is tracked as a [background ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.09907626360654831,
-0.07250998914241791,
-0.12887118756771088,
0.00896446593105793,
-0.029471028596162796,
-0.06030678749084473,
0.017093362286686897,
0.045966796576976776,
0.02457244321703911,
-0.014092572964727879,
0.011478310450911522,
-0.023363059386610985,
-0.0024082977324724197,
0... | 0.147815 |
allowed by `acp.allowedAgents` when that allowlist is set. - The harness command can start on the Gateway host. - Provider auth is present for that harness (`claude`, `codex`, `gemini`, `opencode`, `droid`, etc.). - The selected model exists for that harness - model ids are not portable across harnesses. - The requeste... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.05427255481481552,
-0.032593101263046265,
-0.11653211712837219,
0.03072209283709526,
-0.04467363655567169,
-0.021851854398846626,
-0.038196999579668045,
0.02774413488805294,
-0.06687680631875992,
0.02325134538114071,
0.0730624869465828,
-0.018190573900938034,
0.008331669494509697,
0.037... | 0.054976 |
orphaned parent-owned one-shot ACP sessions. Persistent ACP sessions are preserved while an active conversation binding remains; stale persistent sessions without an active binding are closed so they cannot be silently resumed after the owning task is done or its task record is gone. - Bound follow-up messages go direc... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.09661752730607986,
-0.0033671522978693247,
-0.01361113041639328,
0.07215037196874619,
0.017107339575886726,
-0.008272489532828331,
0.022591132670640945,
-0.041034381836652756,
0.032802991569042206,
0.005541674327105284,
0.05868969485163689,
0.09914082288742065,
-0.009536722674965858,
-0... | 0.064682 |
uses `runtime.acp.agent` as the underlying harness id. ## ACP versus sub-agents Use ACP when you want an external harness runtime. Use \*\*native Codex app-server\*\* for Codex conversation binding/control when the `codex` plugin is enabled. Use \*\*sub-agents\*\* when you want OpenClaw-native delegated runs. | Area | ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.029947126284241676,
-0.029300153255462646,
-0.15159866213798523,
0.033929355442523956,
-0.024319523945450783,
-0.04122614860534668,
0.020845923572778702,
0.02154320292174816,
0.029830515384674072,
-0.0020312380511313677,
0.04757580906152725,
-0.08907410502433777,
0.010269650258123875,
-... | 0.098831 |
ACP session. - ACP output is delivered back to the same thread. - Unfocus/close/archive/idle-timeout or max-age expiry removes the binding. - `/acp close`, `/acp cancel`, `/acp status`, `/status`, and `/unfocus` are Gateway commands, not prompts to the ACP harness. Required feature flags for thread-bound ACP: - `acp.en... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.048221126198768616,
-0.0038938254583626986,
-0.015889346599578857,
0.06946731358766556,
0.017557242885231972,
-0.003003649180755019,
0.016839073970913887,
-0.032238464802503586,
0.029601721093058586,
0.04103974625468254,
0.05986584722995758,
-0.020516470074653625,
-0.04544694349169731,
... | 0.086927 |
an ACP session from an agent turn or tool call. ```json { "task": "Open the repo and summarize failing tests", "runtime": "acp", "agentId": "codex", "thread": true, "mode": "session" } ``` `runtime` defaults to `subagent`, so set `runtime: "acp"` explicitly for ACP sessions. If `agentId` is omitted, OpenClaw uses `acp.... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.10244935005903244,
0.016878200694918633,
-0.10859225690364838,
0.05830914527177811,
-0.045876841992139816,
-0.07452282309532166,
0.004830207675695419,
0.023635243996977806,
0.021225234493613243,
-0.0002148855710402131,
0.026227470487356186,
-0.042511362582445145,
-0.022064119577407837,
... | 0.062 |
No binding. Session starts unbound. | Notes: - On non-thread binding surfaces, default behavior is effectively `off`. - Thread-bound spawn requires channel policy support: - Discord: `channels.discord.threadBindings.spawnSessions=true` - Telegram: `channels.telegram.threadBindings.spawnSessions=true` - Use `--bind here... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.05575353652238846,
-0.05507509782910347,
-0.03428398817777634,
0.04272344335913658,
-0.015487076714634895,
-0.06041940674185753,
0.07967936247587204,
-0.0019608456641435623,
0.05147652328014374,
0.031708840280771255,
0.017730098217725754,
0.03100799396634102,
-0.017366163432598114,
0.04... | 0.094708 |
because the completion path is already responsible for the result. Use `resumeSessionId` to continue a previous ACP session instead of starting fresh. The agent replays its conversation history via `session/load`, so it picks up with full context of what came before. ```json { "task": "Continue where we left off - fix ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.07964631170034409,
0.0036884467117488384,
-0.024438651278614998,
-0.00602990947663784,
0.03455004096031189,
-0.009141483344137669,
0.01910114288330078,
0.025134136900305748,
0.07324613630771637,
-0.029986567795276642,
0.007289514411240816,
0.04451947659254074,
-0.04433141276240349,
0.04... | 0.090141 |
| `/acp spawn` | Create ACP session; optional current bind or thread bind. | `/acp spawn codex --bind here --cwd /repo` | | `/acp cancel` | Cancel in-flight turn for target session. | `/acp cancel agent:codex:acp:` | | `/acp steer` | Send steer instruction to running session. | `/acp steer --session support inbox prior... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.06683488190174103,
0.008967066183686256,
-0.10259248316287994,
0.03907855972647667,
-0.040479011833667755,
0.01557120680809021,
0.044434573501348495,
0.061693623661994934,
-0.049331922084093094,
0.05754275992512703,
0.07420617341995239,
-0.022414464503526688,
-0.020764267072081566,
-0.0... | 0.09297 |
| Set `acp.enabled=true`. | | `ACP dispatch is disabled by policy (acp.dispatch.enabled=false)` | Automatic dispatch from normal thread messages disabled. | Set `acp.dispatch.enabled=true` to resume automatic thread routing; explicit `sessions\_spawn({ runtime: "acp" })` calls still work. | | `ACP agent "" is not allow... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.09106666594743729,
0.006505297496914864,
-0.053786974400281906,
0.025775477290153503,
-0.014548297971487045,
0.020069029182195663,
-0.040601927787065506,
0.03588046133518219,
-0.06672896444797516,
0.03152093663811684,
0.07559943199157715,
0.016457440331578255,
-0.06812933087348938,
0.03... | 0.084806 |
session fails early with little output | Permission prompts are blocked by `permissionMode`/`nonInteractivePermissions`. | Check gateway logs for `AcpRuntimeError`. For full permissions, set `permissionMode=approve-all`; for graceful degradation, set `nonInteractivePermissions=deny`. | | ACP session stalls indefinitely... | https://github.com/openclaw/openclaw/blob/main/docs/tools/acp-agents.md | main | opebclaw | [
-0.04022359848022461,
0.0010661935666576028,
-0.00773487938567996,
0.047164082527160645,
0.005628885235637426,
-0.05203542485833168,
-0.010910955257713795,
-0.016596399247646332,
-0.008847917430102825,
0.020643513649702072,
0.05060550570487976,
-0.03138751909136772,
-0.0622817724943161,
0.... | 0.053933 |
## Manual login (recommended) When a site requires login, \*\*sign in manually\*\* in the \*\*host\*\* browser profile (the openclaw browser). Do \*\*not\*\* give the model your credentials. Automated logins often trigger anti-bot defenses and can lock the account. Back to the main browser docs: [Browser](/tools/browse... | https://github.com/openclaw/openclaw/blob/main/docs/tools/browser-login.md | main | opebclaw | [
-0.08242862671613693,
-0.03590228036046028,
-0.06038261950016022,
0.020210199058055878,
0.03916941210627556,
-0.07365208864212036,
-0.002655918011441827,
0.022427355870604515,
-0.030245058238506317,
-0.05676133558154106,
0.007432329934090376,
-0.06731483340263367,
0.04425394907593727,
0.01... | 0.06054 |
OpenClaw supports \*\*Ollama Web Search\*\* as a bundled `web\_search` provider. It uses Ollama's web-search API and returns structured results with titles, URLs, and snippets. For local or self-hosted Ollama, this setup does not need an API key by default. It does require: - an Ollama host that is reachable from OpenC... | https://github.com/openclaw/openclaw/blob/main/docs/tools/ollama-search.md | main | opebclaw | [
-0.0677989199757576,
0.03601182997226715,
-0.03877098485827446,
0.0529334656894207,
-0.035467728972435,
-0.10326459258794785,
-0.12225699424743652,
-0.025956543162465096,
0.023936064913868904,
-0.07663185894489288,
0.020613273605704308,
-0.004760328680276871,
0.008174348622560501,
-0.01374... | 0.005099 |
OpenClaw can use \*\*Firecrawl\*\* in three ways: - as the `web\_search` provider - as explicit plugin tools: `firecrawl\_search` and `firecrawl\_scrape` - as a fallback extractor for `web\_fetch` It is a hosted extraction/search service that supports bot circumvention and caching, which helps with JS-heavy sites or pa... | https://github.com/openclaw/openclaw/blob/main/docs/tools/firecrawl.md | main | opebclaw | [
-0.11359713226556778,
0.047147661447525024,
-0.055968910455703735,
0.08473730832338333,
0.03063923865556717,
-0.08079367876052856,
-0.08872709423303604,
-0.04495964199304581,
0.03298957273364067,
-0.07120230793952942,
-0.011170311830937862,
0.009353207424283028,
0.001498033874668181,
-0.02... | 0.032989 |
-- all providers and auto-detection - [Web Fetch](/tools/web-fetch) -- web\_fetch tool with Firecrawl fallback - [Tavily](/tools/tavily) -- search + extract tools | https://github.com/openclaw/openclaw/blob/main/docs/tools/firecrawl.md | main | opebclaw | [
-0.11752872169017792,
0.02993682026863098,
-0.0770236998796463,
0.03719755634665489,
0.05553428828716278,
-0.05178903788328171,
-0.011210927739739418,
-0.059573255479335785,
-0.02041345275938511,
-0.0735430121421814,
0.0362061932682991,
0.008504110388457775,
0.0053507741540670395,
-0.04273... | 0.092771 |
# Goal A \*\*goal\*\* is one durable objective attached to the current OpenClaw session. It gives the agent and the operator a shared target for long-running work, without turning that target into a background task, reminder, cron job, or standing order. Goals are session state. They move with the session key, survive ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/goal.md | main | opebclaw | [
-0.05329158157110214,
0.03212520107626915,
-0.0947347953915596,
0.007444967050105333,
0.0494433268904686,
0.02463991567492485,
0.030706634745001793,
0.03849063068628311,
0.024885986000299454,
0.02796497941017151,
-0.017281388863921165,
-0.063921257853508,
0.004755387082695961,
0.0158844403... | 0.126488 |
stored with the goal and measured from the session's fresh token count at creation time. If the current session only has stale or unknown token usage when the goal starts, OpenClaw waits for the next fresh session token snapshot and uses that as the baseline, so tokens spent before the goal existed are not charged to t... | https://github.com/openclaw/openclaw/blob/main/docs/tools/goal.md | main | opebclaw | [
-0.025542665272951126,
0.042331259697675705,
-0.09926223754882812,
0.0917961448431015,
0.06407888233661652,
-0.0642322525382042,
0.04308278113603592,
0.03005220554769039,
0.06935638934373856,
0.047295667231082916,
-0.024236073717474937,
-0.0670437216758728,
0.016413690522313118,
-0.0262741... | 0.102275 |
tool](/concepts/session-tool) - [Compaction](/concepts/compaction) - [Task Flow](/automation/taskflow) - [Standing orders](/automation/standing-orders) | https://github.com/openclaw/openclaw/blob/main/docs/tools/goal.md | main | opebclaw | [
-0.0908304899930954,
0.02496207319200039,
-0.08350766450166702,
-0.0268643107265234,
-0.02719856984913349,
0.01481261383742094,
0.034264322370290756,
0.003951346967369318,
0.011590210720896721,
0.04193165898323059,
0.034314896911382675,
0.06741303950548172,
0.026667125523090363,
-0.0097906... | 0.242323 |
`openclaw agent` runs a single agent turn from the command line without needing an inbound chat message. Use it for scripted workflows, testing, and programmatic delivery. ## Quick start ```bash openclaw agent --agent main --message "What is the weather today?" ``` This sends the message through the Gateway and prints ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/agent-send.md | main | opebclaw | [
-0.00976643431931734,
0.008895796723663807,
-0.08477776497602463,
0.06363419443368912,
-0.01972777023911476,
-0.10122893750667572,
0.03531437739729881,
-0.019153287634253502,
0.033402830362319946,
-0.005911763291805983,
-0.014702084474265575,
-0.06334362179040909,
0.02591579593718052,
0.01... | 0.079812 |
--message "Alert" --deliver --reply-channel telegram --reply-to "@admin" ``` ## Related Full `openclaw agent` flag and option reference. Background sub-agent spawning. How session keys work and how `--to`, `--agent`, and `--session-id` resolve them. Native command catalog used inside agent sessions. | https://github.com/openclaw/openclaw/blob/main/docs/tools/agent-send.md | main | opebclaw | [
-0.012674998492002487,
0.009807194583117962,
-0.07540295273065567,
0.028118301182985306,
0.01604752615094185,
-0.10221357643604279,
0.10204868018627167,
-0.020948132500052452,
0.01198642048984766,
0.013702832162380219,
0.034523673355579376,
-0.07537241280078888,
0.01682104729115963,
0.0255... | 0.147684 |
OpenClaw supports Grok as a `web\_search` provider, using xAI web-grounded responses to produce AI-synthesized answers backed by live search results with citations. Grok web search prefers your existing xAI OAuth sign-in when one is available. If no OAuth profile exists, the same xAI API key can also power the built-in... | https://github.com/openclaw/openclaw/blob/main/docs/tools/grok-search.md | main | opebclaw | [
-0.10657374560832977,
-0.030627606436610222,
-0.10740736126899719,
0.03741725534200668,
0.05741351470351219,
-0.08750155568122864,
-0.02416197769343853,
0.014454103074967861,
0.03507354483008385,
-0.08857376873493195,
0.03954025357961655,
-0.059759482741355896,
0.09335370361804962,
-0.0217... | 0.05768 |
OpenClaw supports Kimi as a `web\_search` provider, using Moonshot web search to produce AI-synthesized answers with citations. ## Get an API key Get an API key from [Moonshot AI](https://platform.moonshot.cn/). Set `KIMI\_API\_KEY` or `MOONSHOT\_API\_KEY` in the Gateway environment, or configure via: ```bash openclaw ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/kimi-search.md | main | opebclaw | [
-0.07416155189275742,
0.006129549816250801,
-0.03518400341272354,
0.07214035093784332,
0.059226639568805695,
-0.0463981032371521,
-0.09409105032682419,
-0.04178144037723541,
-0.02846488170325756,
-0.03716262802481651,
0.012441678903996944,
-0.05458107963204384,
0.06208931654691696,
-0.0050... | 0.073081 |
The `web\_fetch` tool does a plain HTTP GET and extracts readable content (HTML to markdown or text). It does \*\*not\*\* execute JavaScript. For JS-heavy sites or login-protected pages, use the [Web Browser](/tools/browser) instead. ## Quick start `web\_fetch` is \*\*enabled by default\*\* -- no configuration needed. ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/web-fetch.md | main | opebclaw | [
-0.08833101391792297,
0.054155051708221436,
-0.04360629618167877,
0.062232356518507004,
0.039470456540584564,
-0.08515118062496185,
-0.08876045793294907,
-0.017836574465036392,
0.05994170904159546,
-0.04279303923249245,
0.008704835548996925,
0.07053934782743454,
-0.00762982526794076,
-0.02... | 0.063146 |
true`. In this mode, OpenClaw still applies hostname-based SSRF checks before sending the request, but it lets the proxy resolve DNS instead of doing local DNS pinning. Enable this only when the proxy is operator-controlled and enforces outbound policy after DNS resolution. If no HTTP(S) proxy env var is configured, or... | https://github.com/openclaw/openclaw/blob/main/docs/tools/web-fetch.md | main | opebclaw | [
-0.06349199265241623,
-0.04615122079849243,
-0.06395193934440613,
0.03963886946439743,
-0.052336130291223526,
-0.09153184294700623,
-0.07187893986701965,
0.019939852878451347,
-0.004933239426463842,
0.007531289476901293,
0.007485796231776476,
0.03299242630600929,
0.03335152938961983,
0.031... | -0.030197 |
OpenClaw agents can generate videos from text prompts, reference images, or existing videos. Sixteen provider backends are supported, each with different model options, input modes, and feature sets. The agent picks the right provider automatically based on your configuration and available API keys. The `video\_generat... | https://github.com/openclaw/openclaw/blob/main/docs/tools/video-generation.md | main | opebclaw | [
-0.009442171081900597,
-0.0788075253367424,
-0.1213797926902771,
0.02176240272819996,
0.08504980802536011,
-0.047081269323825836,
-0.07813934236764908,
-0.02884606644511223,
0.05203084647655487,
-0.07734869420528412,
0.007468255702406168,
-0.04605476185679436,
0.03905240073800087,
-0.00477... | 0.00851 |
one. Use `action: "status"` to check explicitly without triggering a new generation. ## Supported providers | Provider | Default model | Text | Image ref | Video ref | Auth | | --------------------- | ------------------------------- | :--: | ---------------------------------------------------- | -----------------------... | https://github.com/openclaw/openclaw/blob/main/docs/tools/video-generation.md | main | opebclaw | [
-0.030561210587620735,
-0.07548663765192032,
-0.05788714066147804,
0.008906686678528786,
0.022175457328557968,
0.025823205709457397,
-0.022710274904966354,
-0.05425426363945007,
0.03662228584289551,
0.014445429667830467,
0.0801834911108017,
-0.02254224754869938,
0.010387635789811611,
0.005... | -0.006762 |
`videoToVideo` skipped because this org/input path currently needs provider-side video edit access | | OpenRouter | ✓ | ✓ | - | `generate`, `imageToVideo` | | Qwen | ✓ | ✓ | ✓ | `generate`, `imageToVideo`; `videoToVideo` skipped because this provider needs remote `http(s)` video URLs | | Runway | ✓ | ✓ | ✓ | `generate`... | https://github.com/openclaw/openclaw/blob/main/docs/tools/video-generation.md | main | opebclaw | [
0.042540043592453,
-0.07826148718595505,
-0.09258309751749039,
-0.053980935364961624,
0.10207777470350266,
-0.004997096955776215,
-0.10500016063451767,
-0.0028163185343146324,
0.025030016899108887,
-0.018548613414168358,
0.0367036797106266,
-0.08601769804954529,
-0.014488264918327332,
0.01... | 0.011961 |
when a fallback provider exposes a different control surface. Truly unsupported overrides are ignored on a best-effort basis and reported as warnings in the tool result. Hard capability limits (such as too many reference inputs) fail before submission. Tool results report applied settings; `details.normalization` captu... | https://github.com/openclaw/openclaw/blob/main/docs/tools/video-generation.md | main | opebclaw | [
-0.011391534470021725,
-0.06243188679218292,
-0.03337940201163292,
-0.06043068692088127,
0.01535730343312025,
-0.004947377368807793,
-0.029830459505319595,
0.024948380887508392,
-0.06169193983078003,
-0.058950722217559814,
0.015818530693650246,
-0.04157037287950516,
-0.018334971740841866,
... | 0.039719 |
the [`@openclaw/byteplus-modelark`](https://www.npmjs.com/package/@openclaw/byteplus-modelark) plugin. Provider id: `byteplus-seedance15`. Model: `seedance-1-5-pro-251215`. Uses the unified `content[]` API. Supports at most 2 input images (`first\_frame` + `last\_frame`). All inputs must be remote `https://` URLs. Set ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/video-generation.md | main | opebclaw | [
-0.024096829816699028,
-0.08735240250825882,
-0.042241599410772324,
-0.026031723245978355,
0.07731622457504272,
0.007910370826721191,
-0.078870989382267,
-0.020446719601750374,
-0.035705626010894775,
-0.020809046924114227,
0.004674999509006739,
-0.03471720591187477,
-0.0020730209071189165,
... | 0.062585 |
for every non-FAL provider in the sweep. - One-second lobster prompt. - Per-provider operation cap from `OPENCLAW\_LIVE\_VIDEO\_GENERATION\_TIMEOUT\_MS` (`180000` by default). FAL is opt-in because provider-side queue latency can dominate release time: ```bash pnpm test:live:media video --video-providers fal ``` Set `O... | https://github.com/openclaw/openclaw/blob/main/docs/tools/video-generation.md | main | opebclaw | [
0.04052730277180672,
-0.04358261451125145,
-0.08209541440010071,
-0.04356355592608452,
0.08263353258371353,
-0.0825798436999321,
-0.0820273905992508,
-0.059402380138635635,
0.01973719336092472,
-0.021642128005623817,
0.04854264482855797,
-0.010817031376063824,
-0.09060242772102356,
0.00774... | -0.096257 |
OpenClaw supports DuckDuckGo as a \*\*key-free\*\* `web\_search` provider. No API key or account is required. DuckDuckGo is an \*\*experimental, unofficial\*\* integration that pulls results from DuckDuckGo's non-JavaScript search pages - not an official API. Expect occasional breakage from bot-challenge pages or HTML ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/duckduckgo-search.md | main | opebclaw | [
-0.05641363933682442,
0.025279363617300987,
-0.028195573017001152,
0.06758071482181549,
0.0862189456820488,
-0.05029963701963425,
-0.1051165834069252,
-0.0683872327208519,
-0.003385890508070588,
-0.05046643316745758,
-0.021822631359100342,
-0.04757608100771904,
-0.0033530956134200096,
0.03... | -0.037562 |
[Tavily](https://tavily.com) is a search API designed for AI applications. OpenClaw exposes it in two ways: - as the `web\_search` provider for the generic search tool - as explicit plugin tools: `tavily\_search` and `tavily\_extract` Tavily returns structured results optimized for LLM consumption with configurable sea... | https://github.com/openclaw/openclaw/blob/main/docs/tools/tavily.md | main | opebclaw | [
-0.10695663094520569,
0.0029111152980476618,
-0.06792769581079483,
0.05767660588026047,
0.02919093333184719,
-0.052712805569171906,
-0.058670006692409515,
-0.004032096825540066,
0.0019569413270801306,
-0.05070556700229645,
-0.0023117277305573225,
0.004118569660931826,
0.04701697826385498,
... | 0.146557 |
request). Use `query` plus `chunks\_per\_source` to get only relevant content instead of full pages. ## Choosing the right tool | Need | Tool | | ------------------------------------ | ---------------- | | Quick web search, no special options | `web\_search` | | Search with depth, topic, AI answers | `tavily\_search` |... | https://github.com/openclaw/openclaw/blob/main/docs/tools/tavily.md | main | opebclaw | [
-0.09810364246368408,
-0.003436655504629016,
-0.036207713186740875,
0.05381922796368599,
0.010868065990507603,
-0.05102596431970596,
-0.017970342189073563,
-0.002958515891805291,
-0.02676691859960556,
0.008690859191119671,
0.0448806993663311,
-0.03322577476501465,
0.08027084171772003,
-0.0... | 0.011438 |
Skill Workshop is OpenClaw's governed path for creating and updating workspace skills. Agents and operators do not write active `SKILL.md` files directly through this path. They create a \*\*proposal\*\* first. A proposal is a pending draft containing the proposed skill content, target binding, scanner state, hashes, s... | https://github.com/openclaw/openclaw/blob/main/docs/tools/skill-workshop.md | main | opebclaw | [
-0.10397637635469437,
-0.04941899701952934,
-0.06434150040149689,
0.02709740586578846,
-0.003324744291603565,
0.00021261951769702137,
-0.0024987312499433756,
-0.06535394489765167,
-0.039417605847120285,
0.047069769352674484,
0.010904856957495213,
0.02059505134820938,
0.03879877179861069,
-... | 0.115072 |
absolute paths, hidden path segments, path traversal, overlapping paths, executable files from proposal directories, non-UTF-8 text, null bytes, and files outside the standard support folders. ## Agent tool The model uses `skill\_workshop`: ```text action: create | update | revise | list | inspect | apply | reject | qu... | https://github.com/openclaw/openclaw/blob/main/docs/tools/skill-workshop.md | main | opebclaw | [
-0.0737074688076973,
-0.010504540055990219,
-0.018813416361808777,
0.02937323972582817,
0.02559066005051136,
-0.05739359185099602,
-0.08436509966850281,
-0.039129503071308136,
-0.010221737436950207,
0.026652416214346886,
0.07816217094659805,
-0.004719768650829792,
-0.026045076549053192,
0.... | 0.140149 |
Tool Search is an experimental OpenClaw agent runtime feature. It gives agents one compact way to discover and call large tool catalogs. It is useful when the run has many available tools but the model is likely to need only a few of them. This page documents OpenClaw Tool Search. It is not the Codex-native tool search... | https://github.com/openclaw/openclaw/blob/main/docs/tools/tool-search.md | main | opebclaw | [
-0.09278810024261475,
-0.043721433728933334,
-0.06248181313276291,
0.10298895090818405,
0.021519387140870094,
-0.10428620874881744,
-0.0834415927529335,
0.017720159143209457,
0.03823947533965111,
-0.050283439457416534,
-0.01118698250502348,
0.008828445337712765,
0.008409455418586731,
-0.03... | 0.112741 |
model loads only the tool schemas it actually needs Direct tool exposure is still the right default for small catalogs. Tool Search is best when one run can see many tools, especially from MCP servers or client-provided app tools. ## API `openclaw.tools.search(query, options?)` Searches the effective catalog for the cu... | https://github.com/openclaw/openclaw/blob/main/docs/tools/tool-search.md | main | opebclaw | [
-0.00498623913154006,
0.03318571299314499,
0.018165133893489838,
0.15019749104976654,
0.001148412236943841,
-0.02274067886173725,
-0.10205236077308655,
0.026828061789274216,
0.05984693020582199,
-0.05137576907873154,
-0.06204775720834732,
-0.07618217915296555,
-0.025532741099596024,
0.0021... | 0.036595 |
a tool is not in the effective policy, search should not return it - if a selected tool becomes unavailable, `tool\_call` should fail - if policy or approval blocks execution, the call result should report that block instead of bypassing it - if the code bridge cannot create an isolated runtime, use `mode: "tools"` or ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/tool-search.md | main | opebclaw | [
-0.08153773844242096,
-0.01701211929321289,
-0.05235598608851433,
0.028721068054437637,
-0.017460521310567856,
-0.04814167693257332,
-0.03370477259159088,
-0.024312160909175873,
-0.023464560508728027,
0.02285912074148655,
0.08938570320606232,
-0.05738729238510132,
0.014623571187257767,
-0.... | 0.019012 |
`/steer` first tries to send guidance to an already-active run. It is for "adjust this run while it is still working" moments. If the current runtime cannot accept steering, OpenClaw sends the message as a normal prompt instead of dropping it. ## Current session Use top-level `/steer` to target the active run for the c... | https://github.com/openclaw/openclaw/blob/main/docs/tools/steer.md | main | opebclaw | [
-0.007248672191053629,
-0.04073040932416916,
-0.04550386220216751,
0.08489374816417694,
0.01341093797236681,
-0.10051045566797256,
0.025241609662771225,
-0.0022215493954718113,
-0.009436567313969135,
0.02912532538175583,
0.05207875743508339,
-0.021211013197898865,
0.01941538415849209,
-0.0... | 0.093541 |
Sub-agents are background agent runs spawned from an existing agent run. They run in their own session (`agent::subagent:`) and, when finished, \*\*announce\*\* their result back to the requester chat channel. Each sub-agent run is tracked as a [background task](/automation/tasks). Primary goals: - Parallelize "researc... | https://github.com/openclaw/openclaw/blob/main/docs/tools/subagents.md | main | opebclaw | [
-0.09429530054330826,
-0.029663842171430588,
-0.048089705407619476,
0.018690377473831177,
-0.00794524047523737,
-0.06292268633842468,
0.0013619937235489488,
0.011709293350577354,
0.03946445509791374,
0.012549974024295807,
-0.061820436269044876,
-0.05472683906555176,
0.010461083613336086,
0... | 0.176834 |
- Completion delivery keeps the resolved requester route: thread-bound or conversation-bound completion routes win when available; if the completion origin only provides a channel, OpenClaw fills the missing target/account from the requester session's resolved route (`lastChannel` / `lastTo` / `lastAccountId`) so direc... | https://github.com/openclaw/openclaw/blob/main/docs/tools/subagents.md | main | opebclaw | [
-0.07240341603755951,
-0.04733090102672577,
-0.03179427236318588,
0.03508349880576134,
-0.03162342682480812,
-0.0491824671626091,
-0.0567096471786499,
0.004147619940340519,
0.0176172386854887,
0.08240073174238205,
0.04133501276373863,
-0.014245167374610901,
-0.050571102648973465,
-0.000913... | 0.088624 |
inherit the caller unless you set `agents.defaults.subagents.model` (or per-agent `agents.list[].subagents.model`). ACP runtime spawns use the same configured subagent model when present; otherwise the ACP harness keeps its own default. An explicit `sessions\_spawn.model` still wins. - \*\*Thinking:\*\* native sub-agen... | https://github.com/openclaw/openclaw/blob/main/docs/tools/subagents.md | main | opebclaw | [
-0.04877995699644089,
-0.041618358343839645,
-0.07360545545816422,
0.05914491415023804,
-0.01976732723414898,
-0.03392384946346283,
0.003160958644002676,
0.024891238659620285,
0.01750902272760868,
0.02848244458436966,
0.014818706549704075,
-0.052161868661642075,
0.00402157474309206,
0.0321... | 0.073405 |
later. Target resolution accepts exact `taskName` matches and unambiguous prefixes. Matching is scoped to the same active/recent target window used by numbered `/subagents` targets, so a stale completed child does not make a reused handle ambiguous. If two active or recent children share the same `taskName`, the target... | https://github.com/openclaw/openclaw/blob/main/docs/tools/subagents.md | main | opebclaw | [
-0.06872057169675827,
-0.017939560115337372,
-0.0100545110180974,
0.043968018144369125,
0.027658432722091675,
-0.04130152612924576,
0.04228470101952553,
-0.015311005525290966,
0.03967267647385597,
0.019394848495721817,
-0.008219664916396141,
-0.05407660827040672,
-0.02923567406833172,
0.05... | 0.093393 |
Default: only the requester agent. If you set a list and still want the requester to spawn itself with `agentId`, include the requester id in the list. Default configured target-agent allowlist used when the requester agent does not set its own `subagents.allowAgents`. Block `sessions\_spawn` calls that omit `agentId` ... | https://github.com/openclaw/openclaw/blob/main/docs/tools/subagents.md | main | opebclaw | [
-0.03174673020839691,
-0.010221574455499649,
-0.07435271888971329,
0.023393645882606506,
0.008105084300041199,
-0.033165037631988525,
0.02633790671825409,
-0.022048167884349823,
0.0465734526515007,
0.02087719552218914,
-0.012514452449977398,
-0.005758577957749367,
0.012201285921037197,
-0.... | 0.125819 |
live children remain attached, ended children stay visible for a short recent window, and stale store-only child links are ignored after their freshness window. This prevents old `spawnedBy` / `parentSessionKey` metadata from resurrecting ghost children after restart. If a child completion event arrives after you alrea... | https://github.com/openclaw/openclaw/blob/main/docs/tools/subagents.md | main | opebclaw | [
-0.10249713063240051,
0.012077532708644867,
0.0344666987657547,
0.06449562311172485,
0.014481511898338795,
-0.016853725537657738,
0.01780974492430687,
-0.024802979081869125,
0.025558264926075935,
0.08682194352149963,
0.020267430692911148,
-0.05090402439236641,
-0.012614537961781025,
-0.013... | -0.003975 |
Follow-up | Instruction describing when to reply vs stay silent | Terminal failed runs report failure status without replaying captured reply text. Tool/toolResult output is not promoted into child result text. ### Stats line Announce payloads include a stats line at the end (even when wrapped): - Runtime (e.g. `runtim... | https://github.com/openclaw/openclaw/blob/main/docs/tools/subagents.md | main | opebclaw | [
-0.08860695362091064,
-0.012909071519970894,
-0.06791620701551437,
0.03236658498644829,
-0.04629622772336006,
0.006797327194362879,
-0.009296664036810398,
0.044397059828042984,
0.044616878032684326,
0.028591416776180267,
-0.019647948443889618,
-0.014442238956689835,
0.008043883368372917,
-... | 0.160836 |
on that session and stops auto-resuming it on later restarts. Run `openclaw tasks maintenance --apply` to reconcile the task record, or `openclaw doctor --fix` to clear stale aborted recovery flags on tombstoned sessions. If a sub-agent spawn fails with Gateway `PAIRING\_REQUIRED` / `scope-upgrade`, check the RPC calle... | https://github.com/openclaw/openclaw/blob/main/docs/tools/subagents.md | main | opebclaw | [
-0.08895715326070786,
-0.01628454029560089,
0.0007272085640579462,
0.009277998469769955,
-0.015940766781568527,
-0.054046038538217545,
-0.03014763817191124,
-0.046198584139347076,
-0.0037434010300785303,
0.0007216192316263914,
-0.03556003421545029,
0.06709220260381699,
-0.01753605343401432,
... | 0.037403 |
OpenClaw supports Gemini models with built-in [Google Search grounding](https://ai.google.dev/gemini-api/docs/grounding), which returns AI-synthesized answers backed by live Google Search results with citations. ## Get an API key Go to [Google AI Studio](https://aistudio.google.com/apikey) and create an API key. Set `G... | https://github.com/openclaw/openclaw/blob/main/docs/tools/gemini-search.md | main | opebclaw | [
-0.11512646079063416,
-0.02621150203049183,
0.0046461764723062515,
0.044095057994127274,
0.026826370507478714,
-0.03862149268388748,
-0.09551219642162323,
-0.016764890402555466,
-0.013576354831457138,
-0.03219364210963249,
-0.058244507759809494,
0.003924487624317408,
-0.012319104745984077,
... | -0.002815 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.