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 |
|---|---|---|---|---|---|
uninstall plugins - run doctor repairs that rewrite config or state Applied writes are recorded in: ```text ~/.openclaw/audit/crestodian.jsonl ``` Discovery is not audited. Only applied operations and writes are logged. `openclaw onboard --modern` starts Crestodian as the modern onboarding preview. Plain `openclaw onbo... | https://github.com/openclaw/openclaw/blob/main/docs/cli/crestodian.md | main | opebclaw | [
-0.0405077300965786,
-0.016351673752069473,
-0.020000683143734932,
0.06769069284200668,
0.06384427100419998,
-0.0973944216966629,
-0.03583407774567604,
-0.014399614185094833,
-0.020208586007356644,
0.01332809217274189,
0.04263084754347801,
0.019436217844486237,
-0.01399978343397379,
0.0510... | 0.036752 |
local CLI repair is required. - Default effective state is `auto`: allow remote rescue only in trusted YOLO operation, where the runtime already has unsandboxed local authority. - Require an explicit owner identity. Rescue must not accept wildcard sender rules, open group policy, unauthenticated webhooks, or anonymous ... | https://github.com/openclaw/openclaw/blob/main/docs/cli/crestodian.md | main | opebclaw | [
-0.06843829154968262,
-0.056331563740968704,
0.014729705639183521,
-0.0357939675450325,
0.04903724417090416,
-0.007225933484733105,
-0.049806173890829086,
-0.019397476688027382,
-0.06399624049663544,
0.05054517835378647,
0.05246232822537422,
0.022776823490858078,
-0.03805321082472801,
0.01... | 0.115196 |
# `openclaw cron` Manage cron jobs for the Gateway scheduler. Run `openclaw cron --help` for the full command surface. See [Cron jobs](/automation/cron-jobs) for the conceptual guide. ## Create jobs quickly `openclaw cron create` is an alias for `openclaw cron add`. For new jobs, put the schedule first and the prompt s... | https://github.com/openclaw/openclaw/blob/main/docs/cli/cron.md | main | opebclaw | [
-0.08309558033943176,
0.010443258099257946,
-0.07076795399188995,
0.02436748519539833,
-0.022936563938856125,
-0.1120636984705925,
-0.004666450899094343,
-0.021816903725266457,
-0.021494334563612938,
-0.023461569100618362,
-0.011778869666159153,
-0.09163855016231537,
-0.022144073620438576,
... | 0.060434 |
failure notifications. If an isolated run times out before the first model request, `openclaw cron show` and `openclaw cron runs` include a phase-specific error such as `setup timed out before runner start` or `stalled before first model call (last phase: context-engine)`. For CLI-backed providers, the pre-model watchd... | https://github.com/openclaw/openclaw/blob/main/docs/cli/cron.md | main | opebclaw | [
-0.0688980221748352,
-0.04672393575310707,
-0.031620558351278305,
0.043592825531959534,
0.05691184476017952,
-0.0646132230758667,
-0.06755952537059784,
-0.037028972059488297,
0.03860294073820114,
-0.060651399195194244,
-0.00028483508503995836,
-0.03826988488435745,
0.030591275542974472,
0.... | 0.044402 |
job payload/API) makes the cron run strict. - When a job has `--model` but no fallback list is configured, OpenClaw passes an explicit empty fallback override so the agent primary is not appended as a hidden retry target. - Local-provider preflight checks walk configured fallbacks before marking a cron run `skipped`. `... | https://github.com/openclaw/openclaw/blob/main/docs/cli/cron.md | main | opebclaw | [
-0.08543463051319122,
-0.024390215054154396,
-0.00923466868698597,
0.04661158099770546,
0.008636991493403912,
-0.08323311805725098,
-0.030517978593707085,
-0.006678068544715643,
0.0025535698514431715,
-0.04276137053966522,
0.046191055327653885,
-0.030874285846948624,
-0.020203931257128716,
... | -0.019777 |
``` Create an isolated job with lightweight bootstrap context: ```bash openclaw cron create "0 7 \* \* \*" \ "Summarize overnight updates." \ --name "Lightweight morning brief" \ --session isolated \ --light-context \ --no-deliver ``` `--light-context` applies to isolated agent-turn jobs only. For cron runs, lightweigh... | https://github.com/openclaw/openclaw/blob/main/docs/cli/cron.md | main | opebclaw | [
-0.09401103109121323,
0.014101753942668438,
-0.10186339169740677,
0.10673712939023972,
0.12219998240470886,
-0.07731057703495026,
-0.0037262968253344297,
-0.0009570093243382871,
-0.04047985374927521,
-0.07129942625761032,
0.012440883554518223,
-0.039804257452487946,
0.01711391843855381,
0.... | 0.042736 |
# `openclaw proxy` Validate operator-managed proxy routing, or run the local explicit debug proxy and inspect captured traffic. Use `validate` to preflight an operator-managed forward proxy before enabling OpenClaw proxy routing. The other commands are debugging tools for transport-level investigation: they can start a... | https://github.com/openclaw/openclaw/blob/main/docs/cli/proxy.md | main | opebclaw | [
-0.001616723253391683,
0.01710864156484604,
-0.0373908132314682,
0.012370387092232704,
-0.0374796986579895,
-0.137180358171463,
-0.04253030940890312,
-0.08449900895357132,
-0.05538516119122505,
0.035082973539829254,
-0.007833649404346943,
0.0008785231038928032,
-0.02767168916761875,
0.0110... | -0.006391 |
# `openclaw health` Fetch health from the running Gateway. ## Options | Flag | Default | Description | | ---------------- | ------- | ------------------------------------------------------------------ | | `--json` | `false` | Print machine-readable JSON instead of text. | | `--timeout ` | `10000` | Connection timeout i... | https://github.com/openclaw/openclaw/blob/main/docs/cli/health.md | main | opebclaw | [
0.019235050305724144,
0.053943242877721786,
-0.1294594407081604,
0.038684383034706116,
0.0009000473655760288,
-0.08519986271858215,
-0.031817685812711716,
0.020401768386363983,
-0.008181538432836533,
-0.05291780084371567,
0.0599556639790535,
-0.02004256658256054,
-0.011009677313268185,
-0.... | 0.091221 |
# `openclaw reset` Reset local config/state (keeps the CLI installed). Options: - `--scope `: `config`, `config+creds+sessions`, or `full` - `--yes`: skip confirmation prompts - `--non-interactive`: disable prompts; requires `--scope` and `--yes` - `--dry-run`: print actions without removing files Examples: ```bash ope... | https://github.com/openclaw/openclaw/blob/main/docs/cli/reset.md | main | opebclaw | [
-0.017798859626054764,
0.009828736074268818,
-0.08025661110877991,
0.06051326543092728,
0.022149087861180305,
-0.07898963242769241,
-0.00940899457782507,
-0.05336412787437439,
-0.06557755172252655,
-0.02761312760412693,
0.050580814480781555,
-0.006216390524059534,
0.04032323881983757,
-0.0... | 0.030775 |
# `openclaw security` Security tools (audit + optional fixes). Related: - Security guide: [Security](/gateway/security) ## Audit ```bash openclaw security audit openclaw security audit --deep openclaw security audit --deep --password openclaw security audit --deep --token openclaw security audit --fix openclaw security... | https://github.com/openclaw/openclaw/blob/main/docs/cli/security.md | main | opebclaw | [
-0.06630432605743408,
-0.01497308723628521,
-0.09227091819047928,
0.05921182781457901,
0.056878603994846344,
-0.097154900431633,
0.017797768115997314,
-0.028955277055501938,
0.017688127234578133,
0.01943671517074108,
0.041133467108011246,
-0.017557531595230103,
-0.015882350504398346,
0.040... | 0.038944 |
any enabled `/v1/\*` endpoint). Settings prefixed with `dangerous`/`dangerously` are explicit break-glass operator overrides; enabling one is not, by itself, a security vulnerability report. For the complete dangerous-parameter inventory, see the "Insecure or dangerous flags summary" section in [Security](/gateway/secu... | https://github.com/openclaw/openclaw/blob/main/docs/cli/security.md | main | opebclaw | [
-0.0429588221013546,
0.09708551317453384,
-0.04554663971066475,
0.07728636264801025,
0.08772841840982437,
-0.017828788608312607,
0.006313266232609749,
-0.03645741939544678,
-0.057845115661621094,
-0.04238397628068924,
0.04956173524260521,
0.03802502155303955,
0.03951485827565193,
0.0509428... | 0.093437 |
OpenClaw runs a \*\*single embedded agent runtime\*\* - one agent process per Gateway, with its own workspace, bootstrap files, and session store. This page covers that runtime contract: what the workspace must contain, which files get injected, and how sessions bootstrap against it. ## Workspace (required) OpenClaw us... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/agent.md | main | opebclaw | [
-0.04306557774543762,
-0.012577394023537636,
-0.09515064209699631,
0.04879597947001457,
0.015534660778939724,
-0.08176317811012268,
-0.039113599807024,
-0.025413788855075836,
0.008943754248321056,
-0.028032388538122177,
0.025621136650443077,
0.0204282458871603,
0.03550855442881584,
-0.0056... | 0.065989 |
next LLM call, and no longer skips remaining tool calls from the current assistant message. `/queue steer` is the default active-run behavior. `/queue followup` and `/queue collect` make messages wait for a later turn instead of steering. `/queue interrupt` aborts the active run instead. See [Queue](/concepts/queue) an... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/agent.md | main | opebclaw | [
-0.04900810495018959,
-0.04784116521477699,
-0.021917279809713364,
-0.017566252499818802,
-0.02439657226204872,
-0.07105386257171631,
-0.0233298409730196,
-0.0453098826110363,
0.014474456198513508,
0.01165363471955061,
0.03351627290248871,
-0.019555196166038513,
-0.036832526326179504,
-0.0... | 0.065286 |
## What it is - Pulls provider usage/quota directly from their usage endpoints. - No estimated costs; only provider-reported quota windows or account-state summaries. - Human-readable quota-window status output is normalized to `X% left`, even when an upstream API reports consumed quota, remaining quota, or only raw co... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/usage-tracking.md | main | opebclaw | [
-0.08051365613937378,
0.0472121499478817,
-0.0327586866915226,
0.0267713014036417,
0.040397461503744125,
-0.01932358182966709,
0.11248272657394409,
0.03807135298848152,
0.10308314114809036,
0.04276961460709572,
-0.0409529022872448,
-0.055497489869594574,
-0.03202320635318756,
0.01097327005... | 0.136852 |
`memory\_search` finds relevant notes from your memory files, even when the wording differs from the original text. It works by indexing memory into small chunks and searching them using embeddings, keywords, or both. ## Quick start Memory search uses OpenAI embeddings by default. To use another embedding backend, set ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/memory-search.md | main | opebclaw | [
0.01744682341814041,
-0.01151859387755394,
-0.06086931377649307,
0.06903273612260818,
-0.013410336337983608,
0.017527133226394653,
-0.03416166454553604,
0.03894222900271416,
0.03443052992224693,
-0.0378623865544796,
0.03887559846043587,
0.0640689954161644,
-0.011186989024281502,
0.02128022... | 0.091505 |
}, }, }, }, }, }, } ``` ## Multimodal memory With Gemini Embedding 2, you can index images and audio files alongside Markdown. Search queries remain text, but they match against visual and audio content. See the [Memory configuration reference](/reference/memory-config) for setup. ## Session memory search You can optio... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/memory-search.md | main | opebclaw | [
0.04090610891580582,
-0.046485114842653275,
-0.05514157935976982,
0.05169343575835228,
0.030488086864352226,
0.0050656674429774284,
-0.03819529712200165,
0.005747211165726185,
-0.02250625006854534,
-0.07037166506052017,
0.011186759918928146,
0.02552030421793461,
0.002441276330500841,
-0.00... | 0.072106 |
The private QA stack is meant to exercise OpenClaw in a more realistic, channel-shaped way than a single unit test can. Current pieces: - `extensions/qa-channel`: synthetic message channel with DM, channel, thread, reaction, edit, and delete surfaces. - `extensions/qa-lab`: debugger UI and QA bus for observing the tran... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.09386038035154343,
-0.055501289665699005,
-0.0872654989361763,
0.008016331121325493,
-0.034175824373960495,
-0.054940927773714066,
-0.07747365534305573,
0.024780014529824257,
0.008128225803375244,
0.04072079807519913,
0.05069144070148468,
-0.06040786951780319,
-0.042480651289224625,
0.0... | 0.142965 |
stack with a bind-mounted QA Lab bundle: ```bash pnpm openclaw qa docker-build-image pnpm qa:lab:build pnpm qa:lab:up:fast pnpm qa:lab:watch ``` `qa:lab:up:fast` keeps the Docker services on a prebuilt image and bind-mounts `extensions/qa-lab/web/dist` into the `qa-lab` container. `qa:lab:watch` rebuilds that bundle on... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.03561931848526001,
0.009620602242648602,
-0.024120071902871132,
-0.00874111708253622,
-0.0006733043701387942,
-0.11356920748949051,
-0.04001831263303757,
0.0023409745190292597,
0.06114979833364487,
0.008853282779455185,
0.07891526818275452,
-0.01882227137684822,
-0.012252467684447765,
0... | 0.042009 |
when `MANTIS\_DISCORD\_VIEWER\_CHROME\_PROFILE\_DIR` or `MANTIS\_DISCORD\_VIEWER\_CHROME\_PROFILE\_TGZ\_B64` is configured in the QA environment. That viewer profile is only for visual capture; the pass/fail decision still comes from the Discord REST oracle. CI uses the same command surface in `.github/workflows/qa-liv... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.007457677274942398,
-0.06499481201171875,
-0.08742345869541168,
-0.016208887100219727,
-0.027210019528865814,
-0.06461047381162643,
0.0020359307527542114,
0.02901027351617813,
0.012334571219980717,
0.02036151848733425,
0.034053679555654526,
-0.09069087356328964,
0.04217072203755379,
0.0... | 0.064789 |
JSON verdict and only passes when the model reports positive visible evidence; a negative response that merely quotes the target text fails the assertion. Use `--vision-mode metadata` for a no-model smoke that proves the desktop, browser, screenshot, and video plumbing without calling an image-understanding provider. R... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.01010685320943594,
0.023011332377791405,
-0.12538252770900726,
-0.03769881650805473,
0.0265448447316885,
-0.013982770964503288,
0.010887072421610355,
0.015566603280603886,
0.02171545848250389,
0.019237643107771873,
0.05340342968702316,
-0.04400600120425224,
0.0008832004386931658,
0.1412... | 0.044811 |
present. Keep `--output-dir` under the repo root so the guest can write back through the mounted workspace. ## Telegram, Discord, and Slack QA reference Matrix has a [dedicated page](/concepts/qa-matrix) because of its scenario count and Docker-backed homeserver provisioning. Telegram, Discord, and Slack are smaller - ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.01345167774707079,
-0.0026530465111136436,
-0.07401911169290543,
-0.016830991953611374,
-0.03348756954073906,
-0.00208606687374413,
-0.01444276049733162,
-0.012134608812630177,
0.007684464566409588,
0.04939832165837288,
0.01929866522550583,
-0.05845969542860985,
0.028670169413089752,
0.... | 0.014195 |
started by the child OpenClaw gateway through the bundled Discord plugin. Verifies channel mention handling, that the SUT bot has registered the native `/help` command with Discord, and opt-in Mantis evidence scenarios. Required env when `--credential-source env`: - `OPENCLAW\_QA\_DISCORD\_GUILD\_ID` - `OPENCLAW\_QA\_D... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.02991298772394657,
-0.06684663891792297,
-0.07414854317903519,
0.01887189783155918,
0.010566509328782558,
-0.07232670485973358,
0.014338317327201366,
-0.019450537860393524,
0.015063459984958172,
-0.010012820363044739,
0.07085002213716507,
-0.06765454262495041,
-0.0026077742222696543,
0.... | 0.070483 |
- app-level token (`xapp-...`) of the SUT app with `connections:write`, used by Socket Mode so the SUT app can receive events. Prefer a Slack workspace dedicated to QA over reusing a production workspace. The SUT manifest below intentionally narrows the bundled Slack plugin's production install (`extensions/slack/src/s... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.052463702857494354,
-0.08488056808710098,
-0.02558196522295475,
-0.0037619820795953274,
-0.06398028880357742,
-0.01032270584255457,
-0.020499881356954575,
0.06152423471212387,
0.008847751654684544,
0.04060014337301254,
0.036338672041893005,
-0.009903091937303543,
0.009006974287331104,
0... | 0.067221 |
With `OPENCLAW\_QA\_CONVEX\_SITE\_URL` and `OPENCLAW\_QA\_CONVEX\_SECRET\_MAINTAINER` exported in your shell, register and verify: ```bash pnpm openclaw qa credentials add \ --kind slack \ --payload-file slack-creds.json \ --note "QA Slack pool seed" pnpm openclaw qa credentials list --kind slack --status all --json ``... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.013439369387924671,
-0.047600988298654556,
-0.09573377668857574,
0.02632853016257286,
-0.06900955736637115,
-0.10634586215019226,
-0.06280502676963806,
0.0062950244173407555,
-0.01867944374680519,
0.05596909672021866,
0.03299322351813316,
-0.08908659219741821,
-0.011125638149678707,
0.0... | 0.028206 |
use the pool. Slack payload shape checks currently live in the Slack QA runner rather than the broker; use `{ channelId: string, driverBotToken: string, sutBotToken: string, sutAppToken: string }`, with a Slack channel id like `Cxxxxxxxxxx`. See [Setting up the Slack workspace](#setting-up-the-slack-workspace) for app ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.02933521941304207,
0.009178164415061474,
-0.059728898108005524,
-0.006766265258193016,
-0.11089542508125305,
-0.018735986202955246,
-0.036185212433338165,
0.01867678016424179,
0.03625288978219032,
0.06354807317256927,
0.01788892224431038,
-0.07749319076538086,
0.0388408787548542,
0.0867... | 0.066725 |
- scenario execution - compatibility aliases for older `qa-channel` scenarios Runner plugins own the transport contract: - how `openclaw qa ` is mounted beneath the shared `qa` root - how the gateway is configured for that transport - how readiness is checked - how inbound events are injected - how outbound messages ar... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.08634474128484726,
-0.009552204981446266,
-0.0491902232170105,
0.05410444736480713,
-0.07631377130746841,
-0.03157428652048111,
-0.1047603115439415,
0.025174438953399658,
-0.0017140437848865986,
0.025885313749313354,
-0.019957106560468674,
-0.043679021298885345,
-0.006954112555831671,
0... | 0.105746 |
runs local QA gateway child processes, not Docker. Character eval scenarios should set the persona through `SOUL.md`, then run ordinary user turns such as chat, workspace help, and small file tasks. The candidate model should not be told that it is being evaluated. The command preserves each full transcript, records ba... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-e2e-automation.md | main | opebclaw | [
-0.015188931487500668,
0.006968134548515081,
-0.09609106183052063,
-0.004380115773528814,
0.07757245749235153,
0.00990280881524086,
-0.005210386123508215,
0.035010941326618195,
0.02872348204255104,
-0.006450113374739885,
-0.009708615951240063,
-0.05020705983042717,
-0.033829957246780396,
0... | 0.055136 |
Every model has a context window: the maximum number of tokens it can process. When a conversation approaches that limit, OpenClaw \*\*compacts\*\* older messages into a summary so the chat can continue. ## How it works 1. Older conversation turns are summarized into a compact entry. 2. The summary is saved in the sess... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/compaction.md | main | opebclaw | [
-0.04660753533244133,
-0.013746942393481731,
-0.016008105129003525,
0.05501426383852959,
0.027199765667319298,
-0.05441422760486603,
-0.02883961983025074,
-0.009512454271316528,
0.08739049732685089,
-0.043855514377355576,
-0.011571451090276241,
0.051572978496551514,
-0.02634335309267044,
-... | 0.144321 |
pipeline to create a semantic summary. The byte guard requires `truncateAfterCompaction: true`. Without transcript rotation, the active file would not shrink and the guard remains inactive. ### Successor transcripts When `agents.defaults.compaction.truncateAfterCompaction` is enabled, OpenClaw does not rewrite the exis... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/compaction.md | main | opebclaw | [
-0.0663515031337738,
0.003339914372190833,
-0.031779490411281586,
0.008902167901396751,
0.040176872164011,
-0.04053362458944321,
-0.03785204142332077,
-0.044107429683208466,
0.043723903596401215,
-0.03189224749803543,
0.01182000245898962,
0.07369904220104218,
-0.014677097089588642,
-0.0113... | 0.087979 |
## Overview - A single long-lived \*\*Gateway\*\* owns all messaging surfaces (WhatsApp via Baileys, Telegram via grammY, Slack, Discord, Signal, iMessage, WebChat). - Control-plane clients (macOS app, CLI, web UI, automations) connect to the Gateway over \*\*WebSocket\*\* on the configured bind host (default `127.0.0.... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/architecture.md | main | opebclaw | [
-0.03762947767972946,
-0.01073561329394579,
0.008649113588035107,
-0.02947871759533882,
-0.04078136757016182,
-0.056613970547914505,
0.039214275777339935,
0.010715959593653679,
0.009941940195858479,
0.06475917249917984,
0.0424153134226799,
-0.05125611647963524,
0.04162096604704857,
0.06824... | 0.062773 |
connects still require explicit approval. - Gateway auth (`gateway.auth.\*`) still applies to \*\*all\*\* connections, local or remote. Details: [Gateway protocol](/gateway/protocol), [Pairing](/channels/pairing), [Security](/gateway/security). ## Protocol typing and codegen - TypeBox schemas define the protocol. - JSO... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/architecture.md | main | opebclaw | [
-0.017428111284971237,
-0.0399363674223423,
-0.056007757782936096,
0.033518992364406586,
-0.022784816101193428,
-0.003799419617280364,
-0.013446999713778496,
-0.014444783329963684,
0.003935420420020819,
0.0008500533876940608,
0.014729429967701435,
0.012725367210805416,
-0.010576222091913223,... | 0.048233 |
The Matrix QA lane runs the bundled `@openclaw/matrix` plugin against a disposable Tuwunel homeserver in Docker, with temporary driver, SUT, and observer accounts plus seeded rooms. It is the live transport-real coverage for Matrix. This is maintainer-only tooling. Packaged OpenClaw releases intentionally omit `qa-lab`... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-matrix.md | main | opebclaw | [
-0.011995675042271614,
0.0058157239109277725,
-0.15341074764728546,
0.024595238268375397,
-0.02183038555085659,
-0.056341659277677536,
-0.10693664103746414,
-0.0235228780657053,
-0.028474310413002968,
0.03542586788535118,
0.07352268695831299,
-0.046618495136499405,
-0.013724579475820065,
0... | 0.051646 |
attachment coverage. | | `e2ee-smoke` | Minimum E2EE coverage - basic encrypted reply, thread follow-up, bootstrap success. | | `e2ee-deep` | Exhaustive E2EE state-loss, backup, key, and recovery scenarios. | | `e2ee-cli` | `openclaw matrix encryption setup` and `verify \*` CLI scenarios driven through the QA harness. ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-matrix.md | main | opebclaw | [
-0.04278566315770149,
0.04337336868047714,
-0.07267273962497711,
-0.0026657190173864365,
-0.011130246333777905,
-0.015041683800518513,
-0.046701911836862564,
-0.028385696932673454,
0.0020340532064437866,
0.05404968187212944,
0.07583341747522354,
-0.07451394945383072,
0.008969442918896675,
... | 0.108069 |
`OPENCLAW\_QA\_MATRIX\_DISABLE\_FORCE\_EXIT=1`, expect the process to linger. - \*\*Cleanup error:\*\* look for the printed recovery command (a `docker compose ... down --remove-orphans` invocation) and run it manually to release the homeserver port. - \*\*Flaky negative-assertion windows in CI:\*\* lower `OPENCLAW\_QA... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/qa-matrix.md | main | opebclaw | [
-0.0682055726647377,
0.09636188298463821,
-0.05309238284826279,
0.04333489388227463,
0.031053632497787476,
-0.11385248601436615,
-0.09564809501171112,
-0.0856834277510643,
0.004912044852972031,
0.016117796301841736,
0.029437752440571785,
-0.07667486369609833,
0.01910676248371601,
0.0113630... | 0.055534 |
OpenClaw handles inbound messages through a pipeline of session resolution, queueing, streaming, tool execution, and reasoning visibility. This page maps the path from inbound message to reply. ## Message flow (high level) ``` Inbound message -> routing/bindings -> session key -> queue (if a run is active) -> agent run... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/messages.md | main | opebclaw | [
-0.027042260393500328,
-0.07921525090932846,
-0.03223071247339249,
0.03676401823759079,
0.0249515138566494,
-0.1313295215368271,
0.04636713117361069,
-0.04185810685157776,
0.07878834754228592,
0.015680311247706413,
0.000554367492441088,
0.051953285932540894,
-0.012809866108000278,
-0.00583... | 0.084842 |
they include group messages that did \_not\_ trigger a run (for example, mention-gated messages) and \*\*exclude\*\* messages already in the session transcript. Directive stripping only applies to the \*\*current message\*\* section so history remains intact. Channels that wrap history should set `CommandBody` (or `Raw... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/messages.md | main | opebclaw | [
-0.010023321025073528,
0.006324562709778547,
0.016314515843987465,
0.0418027862906456,
-0.009308621287345886,
-0.021744010969996452,
0.051232025027275085,
-0.0717836394906044,
0.003320083487778902,
0.01770145259797573,
0.040638286620378494,
-0.016727356240153313,
0.011829443275928497,
-0.0... | 0.060073 |
shown only when `/verbose full` is enabled. Defaults live under `agents.defaults.silentReply`; `surfaces..silentReply` can override group/internal policy per surface. Bare silent replies are dropped on all surfaces, so parent sessions stay quiet instead of rewriting sentinel text into fallback chatter. ## Related - [Me... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/messages.md | main | opebclaw | [
0.006415384355932474,
-0.07991984486579895,
0.04535267874598503,
0.07340242713689804,
0.047258246690034866,
-0.06616583466529846,
0.07386117428541183,
-0.08144696801900864,
0.04899066314101219,
0.013185863383114338,
0.031186677515506744,
0.029027849435806274,
-0.03412741795182228,
0.015241... | 0.116849 |
A \*\*context engine\*\* controls how OpenClaw builds model context for each run: which messages to include, how to summarize older history, and how to manage context across subagent boundaries. OpenClaw ships with a built-in `legacy` engine and uses it by default - most users never need to change this. Install and sel... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/context-engine.md | main | opebclaw | [
-0.06663718819618225,
0.011889046989381313,
-0.0666925460100174,
0.06410498917102814,
0.03150380030274391,
-0.06445982307195663,
-0.043534696102142334,
-0.01237670797854662,
-0.025989772751927376,
-0.052491046488285065,
0.02404729835689068,
-0.04789700359106064,
0.007080167531967163,
0.012... | 0.094835 |
{ id: "my-engine", name: "My Context Engine", ownsCompaction: true, }, async ingest({ sessionId, message, isHeartbeat }) { // Store the message in your data store return { ingested: true }; }, async assemble({ sessionId, messages, tokenBudget, availableTools, citationsMode }) { // Return messages that fit the budget re... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/context-engine.md | main | opebclaw | [
-0.1052655354142189,
0.039282411336898804,
-0.08965204656124115,
0.10670982301235199,
-0.01587557978928089,
0.04432990774512291,
0.029320865869522095,
0.003499166574329138,
-0.010107203386723995,
0.0031580524519085884,
0.01911315508186817,
-0.10805106163024902,
0.03208586201071739,
-0.0374... | 0.105915 |
Codex or OpenClaw embedded runtime, or select the legacy context engine.", }, }, } ``` Native Codex and OpenClaw embedded agent runs satisfy `assemble-before-prompt`. Generic CLI backends do not, so engines that require it are rejected before the CLI process starts. ### Failure isolation OpenClaw isolates the selected ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/context-engine.md | main | opebclaw | [
-0.06395461410284042,
0.0018159269820898771,
-0.053688060492277145,
0.040163855999708176,
0.027079813182353973,
-0.099399633705616,
-0.0904117003083229,
0.010953235439956188,
-0.0366315059363842,
-0.07607414573431015,
0.016784613952040672,
-0.04279809072613716,
0.022941499948501587,
0.0110... | 0.075186 |
`openclaw/plugin-sdk/memory-host-core` via `buildActiveMemoryPromptSection(...)`. Trimming old tool results in-memory still runs regardless of which context engine is active. ## Tips - Use `openclaw doctor` to verify your engine is loading correctly. - If switching engines, existing sessions continue with their current... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/context-engine.md | main | opebclaw | [
-0.03937988728284836,
-0.015058920718729496,
-0.048912230879068375,
0.06750588864088058,
0.019356222823262215,
-0.06109188124537468,
-0.08806809782981873,
-0.0035728863440454006,
-0.04238129407167435,
-0.04184730723500252,
0.02904832735657692,
-0.012979768216609955,
-0.04759356379508972,
0... | 0.045038 |
Mantis Slack desktop QA is the real-UI lane for Slack-class bugs that need a Linux desktop, VNC rescue, Slack Web, a real OpenClaw gateway, screenshots, videos, and a PR evidence comment. Use it when unit tests or the headless Slack live lane cannot prove the bug. ## Storage model Mantis uses three different storage la... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/mantis-slack-desktop-runbook.md | main | opebclaw | [
-0.01681005209684372,
-0.019747290760278702,
-0.12579195201396942,
-0.00932848360389471,
-0.03367306664586067,
-0.04607946053147316,
-0.03873758018016815,
-0.01952802576124668,
0.002995810005813837,
0.08623107522726059,
0.07159697264432907,
-0.019245769828557968,
0.015728749334812164,
0.05... | 0.120995 |
when the browser screenshot itself must show Slack Web. ## Hydrate modes | Mode | Use when | Remote behavior | Tradeoff | | ------------- | ----------------------------------------- | ------------------------------------------------------------------------------------- | ------------------------------------------------... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/mantis-slack-desktop-runbook.md | main | opebclaw | [
-0.035999272018671036,
0.008124853484332561,
-0.03578975796699524,
0.01039453037083149,
0.0626688227057457,
-0.04850926250219345,
-0.008736494928598404,
0.033263567835092545,
-0.05264067277312279,
0.030975759029388428,
0.07729858160018921,
-0.06092661991715431,
0.01260419562458992,
0.03176... | 0.032272 |
## Goals - Retry per HTTP request, not per multi-step flow. - Preserve ordering by retrying only the current step. - Avoid duplicating non-idempotent operations. ## Defaults - Attempts: 3 - Max delay cap: 30000 ms - Jitter: 0.1 (10 percent) - Provider defaults: - Telegram min delay: 400 ms - Discord min delay: 500 ms #... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/retry.md | main | opebclaw | [
-0.06826915591955185,
0.0048614381812512875,
0.007114148698747158,
0.06683936715126038,
-0.004480927251279354,
-0.08594834059476852,
-0.052531175315380096,
0.0033654437866061926,
0.030314112082123756,
-0.015195317566394806,
0.022444112226366997,
-0.04167879745364189,
0.042178817093372345,
... | 0.062499 |
The Personal Agent Benchmark Pack is a small repo-backed QA scenario pack for local personal assistant workflows. It is not a generic model benchmark and it does not require a new runner. The pack reuses the private QA stack described in [QA overview](/concepts/qa-e2e-automation), the synthetic [QA channel](/channels/q... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/personal-agent-benchmark-pack.md | main | opebclaw | [
-0.1225428432226181,
0.0031969849951565266,
-0.0436161570250988,
0.0007443345384672284,
-0.010087777860462666,
0.010695071890950203,
0.02116137370467186,
0.04969888925552368,
0.03632147237658501,
0.016617679968476295,
0.050296034663915634,
-0.021021805703639984,
-0.014725863002240658,
0.06... | 0.194554 |
"Context" is \*\*everything OpenClaw sends to the model for a run\*\*. It is bounded by the model's \*\*context window\*\* (token limit). Beginner mental model: - \*\*System prompt\*\* (OpenClaw-built): rules, tools, skills list, time/runtime, and injected workspace files. - \*\*Conversation history\*\*: your messages ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/context.md | main | opebclaw | [
-0.021646292880177498,
-0.013809201307594776,
-0.09064242243766785,
0.06832894682884216,
0.020364228636026382,
-0.08756300806999207,
0.06379873305559158,
0.06444784253835678,
0.010766221210360527,
-0.019291335716843605,
0.035545676946640015,
-0.038805678486824036,
0.06184094771742821,
-0.0... | 0.179177 |
Provider "wrappers" or hidden headers (not visible, still counted). ## How OpenClaw builds the system prompt The system prompt is \*\*OpenClaw-owned\*\* and rebuilt each run. It includes: - Tool list + short descriptions. - Skills list (metadata only; see below). - Workspace location. - Time (UTC + converted user time ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/context.md | main | opebclaw | [
-0.057897668331861496,
0.018174201250076294,
-0.08843820542097092,
0.022540070116519928,
0.08208594471216202,
-0.0716387927532196,
-0.027635280042886734,
-0.010303817689418793,
0.0009336213115602732,
-0.004462661687284708,
0.009372931905090809,
-0.018073152750730515,
0.0685884952545166,
-0... | 0.087233 |
(estimate)` = computed on the fly when no run report exists (or when running via a CLI backend that doesn't generate the report). Either way, it reports sizes and top contributors; it does \*\*not\*\* dump the full system prompt or tool schemas. ## Related Custom context injection via plugins. Summarizing long conversa... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/context.md | main | opebclaw | [
0.04272074997425079,
0.028940251097083092,
-0.033414751291275024,
0.1117088571190834,
0.016610760241746902,
-0.010214319452643394,
0.002928501460701227,
0.0778721421957016,
0.0343388095498085,
0.060882389545440674,
-0.015193448401987553,
-0.08416128903627396,
0.07633785158395767,
-0.057312... | 0.154105 |
An \*\*agent runtime\*\* is the component that owns one prepared model loop: it receives the prompt, drives model output, handles native tool calls, and returns the finished turn to OpenClaw. Runtimes are easy to confuse with providers because both show up near model configuration. They are different layers: | Layer | ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/agent-runtimes.md | main | opebclaw | [
-0.057664062827825546,
-0.04537707567214966,
-0.12167178094387054,
0.08818002045154572,
-0.006132215261459351,
-0.06759409606456757,
0.01155516505241394,
0.009963428601622581,
0.047764915972948074,
-0.04285101965069771,
0.04900320619344711,
-0.020659595727920532,
0.05255982279777527,
-0.03... | 0.106604 |
OpenAI model ref, then asks the Codex app-server runtime to run the embedded agent turn. It does not mean "use API billing," and it does not mean the channel, model provider catalog, or OpenClaw session store becomes Codex. When the bundled `codex` plugin is enabled, natural-language Codex control should use the native... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/agent-runtimes.md | main | opebclaw | [
-0.093209408223629,
-0.05276568606495857,
-0.09958992898464203,
0.02284170500934124,
-0.005479373969137669,
-0.0027524956967681646,
-0.01520150899887085,
0.07011421769857407,
0.008584770374000072,
-0.017047451809048653,
0.0446745939552784,
-0.03594027832150459,
-0.0018193894065916538,
0.01... | 0.155701 |
rewrite unsupported internals. ## Runtime selection OpenClaw chooses an embedded runtime after provider and model resolution: 1. Model-scoped runtime policy wins. This can live in a configured provider model entry or in `agents.defaults.models["provider/model"].agentRuntime` / `agents.list[].models["provider/model"].ag... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/agent-runtimes.md | main | opebclaw | [
-0.040096186101436615,
-0.046750929206609726,
-0.02761610597372055,
0.08349370956420898,
0.013997496105730534,
-0.0343797393143177,
-0.0735013410449028,
-0.01943225786089897,
-0.041409075260162354,
-0.008259574882686138,
0.016855504363775253,
-0.010901135392487049,
0.01624201610684395,
0.0... | 0.032922 |
Determines whether OpenClaw can edit history or only mirror it. | | Do OpenClaw dynamic tools work? | Messaging, sessions, cron, and OpenClaw-owned tools rely on this. | | Do dynamic tool hooks work? | Plugins expect `before\_tool\_call`, `after\_tool\_call`, and middleware around OpenClaw-owned tools. | | Do native to... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/agent-runtimes.md | main | opebclaw | [
-0.07822905480861664,
-0.009779739193618298,
-0.06454934924840927,
0.06444012373685837,
-0.007434047292917967,
-0.07711838185787201,
-0.09654970467090607,
0.011041969992220402,
-0.04513857513666153,
-0.002268370008096099,
-0.005637854803353548,
-0.032054413110017776,
-0.029423760250210762,
... | 0.112368 |
Dreaming is the background memory consolidation system in `memory-core`. It helps OpenClaw move strong short-term signals into durable memory while keeping the process explainable and reviewable. Dreaming is \*\*opt-in\*\* and disabled by default. ## What dreaming writes Dreaming keeps two kinds of output: - \*\*Machin... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/dreaming.md | main | opebclaw | [
0.011369068175554276,
-0.03255736455321312,
-0.03652399033308029,
0.07928863912820816,
0.02423754148185253,
0.006596444174647331,
-0.0524568185210228,
-0.01757330633699894,
0.054967284202575684,
0.018313666805624962,
-0.004691691603511572,
0.10462763160467148,
0.03427935764193535,
-0.08508... | 0.171281 |
can see which staged short-term entries came from historical replay, which promoted items were grounded-led, and clear only grounded-only staged entries without touching ordinary live short-term state. ## Deep ranking signals Deep ranking uses six weighted base signals plus phase reinforcement: | Signal | Weight | Desc... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/dreaming.md | main | opebclaw | [
-0.04140094295144081,
-0.13031333684921265,
0.02108137123286724,
0.02044006995856762,
-0.023764068260788918,
0.061191126704216,
-0.02615457959473133,
0.02590998448431492,
-0.016403092071413994,
-0.0025239307433366776,
-0.07159420847892761,
-0.01366965752094984,
0.04748578742146492,
-0.0417... | 0.090572 |
deep-phase thresholds by default unless overridden with CLI flags. Explain why a specific candidate would or would not promote: ```bash openclaw memory promote-explain "router vlan" openclaw memory promote-explain "router vlan" --json ``` Preview REM reflections, candidate truths, and deep promotion output without writ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/dreaming.md | main | opebclaw | [
0.0038463142700493336,
-0.03759384527802467,
-0.045991282910108566,
0.047494564205408096,
0.030916182324290276,
-0.02276810258626938,
-0.08810847252607346,
0.014495828188955784,
-0.026594916358590126,
-0.03885148838162422,
0.001217721262946725,
-0.030070558190345764,
-0.01564582623541355,
... | 0.085467 |
Session pruning trims \*\*old tool results\*\* from the context before each LLM call. It reduces context bloat from accumulated tool outputs (exec results, file reads, search results) without rewriting normal conversation text. Pruning is in-memory only -- it does not modify the on-disk session transcript. Your full hi... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/session-pruning.md | main | opebclaw | [
-0.005929588805884123,
0.08076803386211395,
-0.018982719630002975,
-0.016589360311627388,
-0.0316544808447361,
-0.04160835221409798,
0.04070104658603668,
0.05420095846056938,
0.044849250465631485,
0.031937502324581146,
0.014342200942337513,
0.06988093256950378,
-0.007651356048882008,
-0.03... | 0.041916 |
Active memory is an optional plugin-owned blocking memory sub-agent that runs before the main reply for eligible conversational sessions. It exists because most memory systems are capable but reactive. They rely on the main agent to decide when to search memory, or on the user to say things like "remember this" or "sea... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/active-memory.md | main | opebclaw | [
-0.05707360804080963,
0.0008403066894970834,
-0.08955274522304535,
0.16314998269081116,
0.015489358454942703,
-0.035587262362241745,
0.045461758971214294,
0.007166875526309013,
0.0752442255616188,
0.012042214162647724,
0.01439516618847847,
0.08214825391769409,
-0.057728782296180725,
0.0235... | 0.164906 |
global form: ```text /active-memory status --global /active-memory off --global /active-memory on --global ``` The global form writes `plugins.entries.active-memory.config.enabled`. It leaves `plugins.entries.active-memory.enabled` on so the command remains available to turn active memory back on later. If you want to ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/active-memory.md | main | opebclaw | [
0.02471907250583172,
-0.028874307870864868,
-0.1410861760377884,
0.16523335874080658,
0.03141416609287262,
-0.011870410293340683,
0.05707088112831116,
0.02008790522813797,
-0.014527508057653904,
0.01494048722088337,
0.026083186268806458,
0.021313317120075226,
-0.0044699315913021564,
-0.044... | 0.100549 |
allowedChatIds: ["ou\_operator\_open\_id", "oc\_small\_ops\_group"], deniedChatIds: ["oc\_large\_public\_group"] ``` ## Where it runs Active memory is a conversational enrichment feature, not a platform-wide inference feature. | Surface | Runs active memory? | | ---------------------------------------------------------... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/active-memory.md | main | opebclaw | [
-0.04338434338569641,
-0.002829753328114748,
-0.07293947786092758,
0.07967419922351837,
0.00808312464505434,
-0.058361805975437164,
0.10098500549793243,
-0.033560942858457565,
0.02456735074520111,
-0.004823954310268164,
0.08412336558103561,
-0.04400179162621498,
-0.009514334611594677,
0.03... | 0.089816 |
- `preference-only`: optimized for favorites, habits, routines, taste, and recurring personal facts Default mapping when `config.promptStyle` is unset: ```text message -> strict recent -> balanced full -> contextual ``` If you set `config.promptStyle` explicitly, that override wins. Example: ```json5 promptStyle: "pref... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/active-memory.md | main | opebclaw | [
-0.039907243102788925,
-0.06730939447879791,
-0.004566285293549299,
0.08725964277982712,
0.02197141759097576,
0.05609843507409096,
-0.02395348809659481,
0.017741704359650612,
-0.07136212289333344,
0.019490696489810944,
0.047983549535274506,
0.04431137815117836,
-0.009920409880578518,
-0.01... | 0.062628 |
`config.thinking` can override the blocking memory sub-agent thinking level: ```json5 thinking: "medium" ``` Default: ```json5 thinking: "off" ``` Do not enable this by default. Active Memory runs in the reply path, so extra thinking time directly increases user-visible latency. `config.promptAppend` adds extra operato... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/active-memory.md | main | opebclaw | [
-0.002335962373763323,
-0.030969349667429924,
-0.07040135562419891,
0.05893189087510109,
-0.022754626348614693,
-0.05542708933353424,
0.027986185625195503,
0.05076431855559349,
0.004831300117075443,
0.01384173147380352,
0.007051506079733372,
0.01982055976986885,
-0.019144713878631592,
0.01... | 0.086786 |
Advanced thinking override for the blocking memory sub-agent; default `off` for speed | | `config.promptOverride` | `string` | Advanced full prompt replacement; not recommended for normal use | | `config.promptAppend` | `string` | Advanced extra instructions appended to the default or overridden prompt | | `config.time... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/active-memory.md | main | opebclaw | [
0.03230126574635506,
-0.024617530405521393,
-0.10762570798397064,
0.05840189382433891,
-0.005297355353832245,
0.07065101712942123,
-0.034110598266124725,
0.05823415890336037,
-0.0473850816488266,
0.004934287164360285,
0.04222474992275238,
0.028813963755965233,
0.011611304245889187,
-0.0168... | 0.05749 |
blocking prompt-build hook budget by default and move cold-start setup grace behind explicit `setupGraceTimeoutMs` config, so the plugin no longer silently extends 15000 ms configs to 45000 ms on the main lane."\_ The embedded recall runner uses the same effective timeout budget, so `setupGraceTimeoutMs` covers both th... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/active-memory.md | main | opebclaw | [
0.002640038961544633,
-0.010880798101425171,
-0.06767771393060684,
0.12186974287033081,
0.029433151707053185,
0.008320675231516361,
-0.05295754596590996,
-0.007333282381296158,
-0.004569396376609802,
-0.0027342974208295345,
0.01930955983698368,
0.010399756953120232,
-0.0187522079795599,
-0... | 0.033165 |
When a normal prompt arrives while a session run is already streaming, OpenClaw tries to send that prompt into the active runtime by default when the queue mode is `steer`. No config entry and no queue directive are required for that default behavior. OpenClaw and the native Codex app-server harness implement the deliv... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/queue-steering.md | main | opebclaw | [
-0.014825432561337948,
-0.07142297178506851,
-0.056175876408815384,
0.052424971014261246,
-0.02216700091958046,
-0.1161840409040451,
-0.012019135989248753,
-0.02373606152832508,
0.03053528442978859,
-0.028929155319929123,
0.03638509288430214,
-0.052364449948072433,
0.0029929603915661573,
-... | 0.053362 |
quiet window before sending batched `turn/steer`. For OpenClaw, active steering itself does not use the debounce timer because OpenClaw naturally batches messages until the next model boundary. ## Related - [Command queue](/concepts/queue) - [Steer](/tools/steer) - [Messages](/concepts/messages) - [Agent loop](/concept... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/queue-steering.md | main | opebclaw | [
0.03140481561422348,
-0.09782474488019943,
-0.053863562643527985,
0.09126874804496765,
0.008958544582128525,
-0.12092388421297073,
0.031516239047050476,
-0.03920391947031021,
-0.0022075138986110687,
-0.0585038848221302,
0.07959678769111633,
-0.002916796365752816,
0.0030124264303594828,
-0.... | 0.040585 |
OpenClaw supports "subscription auth" via OAuth for providers that offer it (notably \*\*OpenAI Codex (ChatGPT OAuth)\*\*). For Anthropic, the practical split is now: - \*\*Anthropic API key\*\*: normal Anthropic API billing - \*\*Anthropic Claude CLI / subscription auth inside OpenClaw\*\*: Anthropic staff told us thi... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/oauth.md | main | opebclaw | [
-0.1039283499121666,
0.02396375499665737,
-0.10110543668270111,
-0.0071807256899774075,
0.006520713679492474,
-0.09882217645645142,
0.017763299867510796,
0.012560594826936722,
0.029592245817184448,
-0.032473064959049225,
-0.01105688139796257,
-0.058707285672426224,
0.02636277675628662,
0.0... | 0.065261 |
/ GLM Coding Plan](/providers/zai). OpenClaw also exposes Anthropic setup-token as a supported token-auth path, but it now prefers Claude CLI reuse and `claude -p` when available. ## Anthropic Claude CLI migration OpenClaw supports Anthropic Claude CLI reuse again. If you already have a local Claude login on the host, ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/oauth.md | main | opebclaw | [
-0.06850617378950119,
-0.010989880189299583,
-0.11096048355102539,
0.005304902791976929,
-0.009497559629380703,
-0.11761745065450668,
-0.011457131244242191,
-0.00019319452985655516,
0.027404673397541046,
-0.002333464100956917,
0.012878748588263988,
-0.021824469789862633,
0.016269413754343987... | 0.057249 |
The workspace is the agent's home. It is the only working directory used for file tools and for workspace context. Keep it private and treat it as memory. This is separate from `~/.openclaw/`, which stores config, credentials, and sessions. The workspace is the \*\*default cwd\*\*, not a hard sandbox. Tools resolve rel... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/agent-workspace.md | main | opebclaw | [
0.016980404034256935,
-0.04012051969766617,
-0.11042128503322601,
0.06068979576230049,
0.017269255593419075,
-0.048363517969846725,
-0.023581847548484802,
-0.040206506848335266,
0.011111441999673843,
-0.04994574934244156,
0.030576981604099274,
0.00047077611088752747,
0.04726315662264824,
0... | 0.04584 |
`agents.defaults.bootstrapTotalMaxChars` (default: 60000). `openclaw setup` can recreate missing defaults without overwriting existing files. ## What is NOT in the workspace These live under `~/.openclaw/` and should NOT be committed to the workspace repo: - `~/.openclaw/openclaw.json` (config) - `~/.openclaw/agents//a... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/agent-workspace.md | main | opebclaw | [
-0.05788261815905571,
-0.035790812224149704,
-0.10465773195028305,
0.04079868644475937,
0.023765264078974724,
-0.06622909009456635,
-0.06184448301792145,
-0.04184572771191597,
-0.03610577806830406,
-0.03687595576047897,
0.055903419852256775,
-0.0017076324438676238,
0.018119098618626595,
0.... | -0.017999 |
This page is the target design for replacing scattered channel inbound, reply dispatch, preview streaming, and outbound delivery helpers with one durable message lifecycle. The short version: - The core primitives should be \*\*receive\*\* and \*\*send\*\*, not \*\*reply\*\*. - A reply is only a relation on an outbound... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.055829692631959915,
-0.039394211024045944,
0.04393979161977768,
0.004585292190313339,
0.022276705130934715,
-0.07447151094675064,
-0.01626899279654026,
0.041587136685848236,
-0.055169299244880676,
0.01909998431801796,
-0.054987020790576935,
0.005442832596600056,
-0.03578803315758705,
0.... | 0.065551 |
adapter for dedupe, locks, queues, and persistence OpenClaw should borrow the vocabulary, not copy the surface. What OpenClaw needs beyond that model: - Durable outbound send intents before direct transport calls. - Explicit send contexts with begin, commit, and fail. - Receive contexts that know platform ack policy. -... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.12774400413036346,
-0.04639632999897003,
-0.0391131266951561,
0.041537992656230927,
0.05537339299917221,
-0.12825286388397217,
-0.035127390176057816,
0.020882904529571533,
0.0548236258327961,
0.020750124007463455,
0.025762036442756653,
-0.016446448862552643,
-0.020113926380872726,
-0.00... | 0.104476 |
intent to future edit, delete, preview finalization, duplicate suppression, and recovery. A receipt can describe one platform message or a multi-part delivery. Chunked text, media plus text, voice plus text, and card fallbacks must preserve all platform ids while still exposing a primary id for threading and later edit... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.1080385148525238,
0.0443454310297966,
-0.012825928628444672,
0.011164124123752117,
0.03811334818601608,
0.0021839109249413013,
0.10154837369918823,
0.04526445269584656,
0.07741150259971619,
-0.0019062866922467947,
0.004752050153911114,
-0.04347483813762665,
-0.03700840100646019,
0.01215... | 0.172147 |
if (ctx.preview?.canFinalizeInPlace) { return await ctx.edit(ctx.preview.receipt, rendered); } return await ctx.send(rendered); }); ``` The helper expands to: ```text begin durable intent -> render -> optional preview/edit/stream work -> mark sending -> final platform send or final edit -> mark committing with raw rece... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.036980580538511276,
-0.022177156060934067,
-0.02046796679496765,
0.0005018594674766064,
0.05884777009487152,
-0.027445100247859955,
-0.03746823966503143,
0.017807159572839737,
0.026759013533592224,
0.013344962149858475,
0.0259123295545578,
-0.037167806178331375,
0.020869577303528786,
0.... | 0.080238 |
state: LiveMessageState, final: RenderedMessageBatch, ): Promise; cancel?( ctx: MessageSendContext, state: LiveMessageState, reason: LiveCancelReason, ): Promise; }; ``` Live state is durable enough to recover or suppress duplicates: ```typescript type LiveMessageState = { mode: "partial" | "block" | "progress" | "nati... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.08662521094083786,
0.011736258864402771,
0.058318838477134705,
0.052217189222574234,
0.03239137679338455,
0.0667470395565033,
0.006136229727417231,
-0.011094791814684868,
0.026186078786849976,
0.058353859931230545,
-0.004894666373729706,
-0.027880145236849785,
-0.05666857957839966,
0.04... | 0.090072 |
migration, generic durable delivery is opt-in for any channel whose existing delivery callback has side effects beyond "send this payload". Legacy entry points are non-durable by default: - `channel.inbound.run` and `dispatchChannelInboundReply` use the channel's delivery callback unless that channel explicitly supplie... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.058184631168842316,
-0.06643743813037872,
0.022691627964377403,
0.04090264439582825,
-0.026743056252598763,
-0.042191844433546066,
-0.058400463312864304,
0.05993099883198738,
-0.1226561963558197,
-0.006968169007450342,
-0.03889590874314308,
0.04349016398191452,
0.013768978416919708,
0.0... | 0.047901 |
must be dropped before `allowBots` authorization. Channels must not implement this with visible-text prefix filters except as a short emergency stopgap; the durable contract is structured origin metadata. ## Internal storage The durable queue should store message send intents, not reply payloads. ```typescript type Dur... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.010811193846166134,
0.03835543617606163,
0.01735762320458889,
0.019785651937127113,
-0.028874218463897705,
0.021039485931396484,
0.030587153509259224,
0.02862444333732128,
-0.00008855979103827849,
0.05611206218600273,
0.003309052437543869,
-0.04672815650701523,
-0.003483004169538617,
0.... | 0.058187 |
final suppression. Send adapter owns comments, topic sessions, media, and voice suppression. | | QQ Bot | Live adapter owns C2C streaming, accumulator timeout, and fallback final send. Render adapter owns media tags and text-as-voice. | | Signal | Simple receive plus send adapter. No live adapter unless signal-cli adds... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.08246871083974838,
-0.03937891125679016,
0.030141988769173622,
-0.0005808106507174671,
0.04438965395092964,
-0.022050419822335243,
0.007071511819958687,
-0.0654735192656517,
0.09002696722745895,
0.0475873202085495,
0.033889178186655045,
-0.026281161233782768,
0.060425300151109695,
-0.00... | 0.099596 |
generic send path preserves the old channel delivery semantics. ### Phase 4: Prepared Dispatcher Bridge - Replace `deliverDurableInboundReplyPayload` with a send-context bridge. - Keep the old helper as a wrapper. - Port Telegram, WhatsApp, Slack, Signal, iMessage, and Discord first because they already have durable-fi... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.06006152927875519,
-0.059180594980716705,
0.0711473897099495,
0.03268707916140556,
0.014322313480079174,
-0.03844800591468811,
-0.009742842055857182,
0.016219792887568474,
-0.004837814252823591,
0.07364294677972794,
-0.00749405799433589,
0.024162042886018753,
-0.010666543617844582,
0.07... | 0.090261 |
redacted when media/error/reply-target mismatch requires normal delivery. - Block streaming and preview streaming do not both deliver the same text. - Media streamed early is not duplicated in final delivery. Channel tests: - Telegram topic reply with polling ack delayed until the receive context's safe completed water... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.048987265676259995,
-0.07202636450529099,
0.04517129436135292,
0.03284797444939613,
0.04972457513213158,
-0.03951163962483406,
-0.012042413465678692,
0.011358905583620071,
0.01225566491484642,
0.03223665803670883,
0.025598205626010895,
0.013075881637632847,
0.011856078170239925,
0.01881... | 0.120958 |
- Reply-named SDK helpers are compatibility exports, not the recommended path. - Durable recovery can replay pending final sends after restart without losing the final response or duplicating already committed sends; sends whose platform outcome is unknown are reconciled before replay or documented as at-least-once for... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/message-lifecycle-refactor.md | main | opebclaw | [
-0.08266164362430573,
-0.04323068633675575,
0.01747448742389679,
0.038844913244247437,
0.08079798519611359,
-0.04390513151884079,
-0.055391233414411545,
0.015250974334776402,
-0.0500505231320858,
0.0015925017651170492,
0.01835663989186287,
0.02009025774896145,
0.031082261353731155,
0.00372... | 0.099043 |
[QMD](https://github.com/tobi/qmd) is a local-first search sidecar that runs alongside OpenClaw. It combines BM25, vector search, and reranking in a single binary, and can index content beyond your workspace memory files. ## What it adds over builtin - \*\*Reranking and query expansion\*\* for better recall. - \*\*Inde... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/memory-qmd.md | main | opebclaw | [
-0.04801240935921669,
-0.03382709249854088,
-0.08070012181997299,
0.03625676408410072,
0.0053429026156663895,
0.014731034636497498,
-0.08247243613004684,
-0.0019554682075977325,
-0.046313654631376266,
0.02707073464989662,
0.006567292846739292,
0.0021662211511284113,
0.04326610267162323,
0.... | 0.126022 |
advertises support for multiple collection filters, OpenClaw searches all same-source collections with one command: ```bash qmd search "router notes" --json -n 10 -c memory-root-main -c memory-dir-main ``` This avoids starting one QMD subprocess for every durable-memory collection. Session transcript collections stay i... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/memory-qmd.md | main | opebclaw | [
-0.0385463573038578,
-0.01125361304730177,
-0.08561373502016068,
0.00516950897872448,
-0.018174605444073677,
-0.0034381805453449488,
-0.10428959876298904,
-0.022045960649847984,
0.050537191331386566,
-0.06192624196410179,
0.004912949167191982,
-0.03458695858716965,
0.014520697295665741,
0.... | -0.031868 |
searches only when the installed QMD advertises support for multiple `-c` filters; otherwise it keeps the older per-collection fallback for correctness. \*\*BM25-only QMD still trying to build llama.cpp?\*\* Set `memory.qmd.searchMode = "search"`. OpenClaw treats that mode as lexical-only, does not run QMD vector statu... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/memory-qmd.md | main | opebclaw | [
-0.014850675128400326,
-0.03328267112374306,
-0.05036734417080879,
0.0007249011541716754,
-0.026043890044093132,
-0.03595961257815361,
-0.10641726106405258,
-0.045478951185941696,
0.02634204365313053,
0.04453584924340248,
0.006856976076960564,
-0.05107153207063675,
0.019291670992970467,
0.... | 0.049188 |
Commitments are short-lived follow-up memories. When enabled, OpenClaw can notice that a conversation created a future check-in opportunity and remember to bring it back later. Examples: - You mention an interview tomorrow. OpenClaw may check in afterward. - You say you are exhausted. OpenClaw may ask later whether you... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/commitments.md | main | opebclaw | [
-0.010923414491117,
0.0017989998450502753,
-0.08278286457061768,
0.05397637188434601,
0.021685689687728882,
-0.0595569908618927,
-0.023813396692276,
-0.04033876955509186,
0.05059156194329262,
-0.047269321978092194,
-0.02173738367855549,
0.02503299154341221,
-0.0002220127935288474,
0.009462... | 0.07555 |
commitments --all openclaw commitments --agent main openclaw commitments --status snoozed openclaw commitments dismiss cm\_abc123 ``` See [`openclaw commitments`](/cli/commitments) for the command reference. ## Privacy and cost Commitment extraction uses an LLM pass, so enabling it adds background model usage after eli... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/commitments.md | main | opebclaw | [
0.010513744316995144,
0.010470292530953884,
-0.09572107344865799,
0.050866641104221344,
0.03232141584157944,
-0.037068888545036316,
-0.003595980815589428,
-0.040868695825338364,
-0.028421541675925255,
0.004495619330555201,
0.012172537855803967,
-0.06135761737823486,
-0.011137116700410843,
... | 0.089388 |
Run multiple \_isolated\_ agents — each with its own workspace, state directory (`agentDir`), and session history — plus multiple channel accounts (e.g. two WhatsApps) in one running Gateway. Inbound messages are routed to the right agent through bindings. An \*\*agent\*\* here is the full per-persona scope: workspace ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/multi-agent.md | main | opebclaw | [
-0.005184129811823368,
-0.08458998054265976,
-0.10370591282844543,
-0.05279011279344559,
-0.054205358028411865,
-0.018172966316342354,
0.06651472300291061,
-0.01452636905014515,
0.03327179327607155,
0.004314872901886702,
0.03589392453432083,
-0.030101293697953224,
0.03656930848956108,
0.02... | 0.111172 |
--account work ``` See channel guides: [Discord](/channels/discord), [Telegram](/channels/telegram), [WhatsApp](/channels/whatsapp). Add agents under `agents.list`, channel accounts under `channels..accounts`, and connect them with `bindings` (examples below). ```bash openclaw gateway restart openclaw agents list --bin... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/multi-agent.md | main | opebclaw | [
-0.03156781196594238,
-0.08636369556188583,
-0.10374758392572403,
-0.029035720974206924,
-0.06755084544420242,
-0.03474437817931175,
0.07746145129203796,
-0.02178795076906681,
0.036377351731061935,
-0.006082165986299515,
0.025943219661712646,
-0.05566246062517166,
0.045809000730514526,
-0.... | 0.13039 |
identify each login. Each `accountId` can be routed to a different agent, so one server can host multiple phone numbers without mixing sessions. If you want a channel-wide default account when `accountId` is omitted, set `channels..defaultAccount` (optional). When unset, OpenClaw falls back to `default` if present, oth... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/multi-agent.md | main | opebclaw | [
-0.02360881119966507,
-0.08456512540578842,
-0.0728965625166893,
-0.020975392311811447,
-0.08635029941797256,
-0.06641971319913864,
0.08345770090818405,
-0.02395668812096119,
0.08206376433372498,
0.006198117975145578,
0.002333632670342922,
-0.027883626520633698,
0.06123786419630051,
0.0188... | 0.033626 |
"whatsapp", accountId: "personal", peer: { kind: "group", id: "1203630...@g.us" }, }, }, ], // Off by default: agent-to-agent messaging must be explicitly enabled + allowlisted. tools: { agentToAgent: { enabled: false, allow: ["home", "work"], }, }, channels: { whatsapp: { accounts: { personal: { // Optional override. ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/multi-agent.md | main | opebclaw | [
-0.0213003046810627,
0.006448879837989807,
-0.044570691883563995,
0.007571185007691383,
-0.001995200291275978,
-0.04389210790395737,
0.018566830083727837,
-0.04798869043588638,
-0.017689649015665054,
0.029039738699793816,
0.007663271389901638,
-0.004422059282660484,
-0.0524660125374794,
0.... | -0.007754 |
untrusted agents. - \*\*Resource control\*\*: sandbox specific agents while keeping others on host. - \*\*Flexible policies\*\*: different permissions per agent. `tools.elevated` is \*\*global\*\* and sender-based; it is not configurable per agent. If you need per-agent boundaries, use `agents.list[].tools` to deny `ex... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/multi-agent.md | main | opebclaw | [
0.006604544352740049,
-0.024150287732481956,
-0.10538225620985031,
0.007847457192838192,
0.049213066697120667,
-0.05495205149054527,
0.06688901782035828,
-0.041519831866025925,
-0.03820722550153732,
-0.004982128739356995,
0.07387492060661316,
-0.0863642618060112,
0.08144465833902359,
0.013... | 0.131558 |
OpenClaw standardizes timestamps so the model sees a \*\*single reference time\*\* instead of a mix of provider-local clocks. There are three surfaces where timezones show up, each with its own purpose: ## Three timezone surfaces | Surface | What it shows | Default | Configured via | | ----------------- | -------------... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/timezone.md | main | opebclaw | [
-0.008992566727101803,
0.004476732108741999,
0.01244052778929472,
0.05759609490633011,
0.02348148822784424,
-0.07250680774450302,
-0.08835891634225845,
-0.04845808446407318,
0.03602641448378563,
-0.03655816987156868,
0.0033662591595202684,
-0.08228399604558945,
-0.07549669593572617,
-0.049... | -0.03649 |
OpenClaw has two separate streaming layers: - \*\*Block streaming (channels):\*\* emit completed \*\*blocks\*\* as the assistant writes. These are normal channel messages (not token deltas). - \*\*Preview streaming (Telegram/Discord/Slack):\*\* update a temporary \*\*preview message\*\* while generating. There is \*\*n... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/streaming.md | main | opebclaw | [
-0.063416488468647,
-0.06900328397750854,
0.014167899265885353,
0.05101079121232033,
0.05596856400370598,
-0.11933590471744537,
-0.04190046340227127,
-0.02154931239783764,
0.08031101524829865,
-0.024510882794857025,
-0.012412318028509617,
-0.06568755954504013,
-0.04034112021327019,
-0.0284... | 0.134591 |
streaming is enabled, you can add a \*\*randomized pause\*\* between block replies (after the first block). This makes multi-bubble responses feel more natural. - Config: `agents.defaults.humanDelay` (override per agent via `agents.list[].humanDelay`). - Modes: `off` (default), `natural` (800-2500ms), `custom` (`minMs`... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/streaming.md | main | opebclaw | [
-0.05408916994929314,
-0.053931184113025665,
0.010451304726302624,
-0.02366834692656994,
0.015846431255340576,
-0.05892851576209068,
0.03245290741324425,
-0.07136442512273788,
0.0047608050517737865,
-0.006022738292813301,
-0.05608474463224411,
-0.09434479475021362,
-0.058365777134895325,
0... | 0.079767 |
stream` can write reasoning to a transient preview that is deleted after final delivery. Discord: - Uses send + edit preview messages. - `block` mode uses draft chunking (`draftChunk`). - Preview streaming is skipped when Discord block streaming is explicitly enabled. - Final media, error, and explicit-reply payloads c... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/streaming.md | main | opebclaw | [
-0.026345867663621902,
-0.0899185985326767,
0.036862198263406754,
0.022377733141183853,
0.06757090985774994,
-0.011906302534043789,
-0.03763142600655556,
-0.0027801005635410547,
0.03790748491883278,
0.07040364295244217,
-0.04711709916591644,
0.020471680909395218,
-0.045860495418310165,
0.0... | 0.074196 |
while hiding command/exec text, set `streaming.preview.commandText` to `"status"` or `streaming.progress.commandText` to `"status"`; the default is `"raw"` to preserve released behavior. This policy is shared by draft/progress channels that use OpenClaw's compact progress renderer, including Discord, Matrix, Microsoft ... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/streaming.md | main | opebclaw | [
-0.048426076769828796,
0.039306409657001495,
-0.019014231860637665,
0.10767284780740738,
0.059363119304180145,
-0.04402711242437363,
-0.032110776752233505,
-0.0326228104531765,
0.059570685029029846,
-0.0008226932259276509,
-0.0021229300182312727,
-0.026831673458218575,
-0.04455925524234772,
... | 0.025875 |
Reference for \*\*LLM/model providers\*\* (not chat channels like WhatsApp/Telegram). For model selection rules, see [Models](/concepts/models). ## Quick rules - Model refs use `provider/model` (example: `opencode/claude-opus-4-6`). - `agents.defaults.models` acts as an allowlist when set. - CLI helpers: `openclaw onbo... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/model-providers.md | main | opebclaw | [
-0.04501627758145332,
-0.09283894300460815,
-0.018998341634869576,
0.03372735530138016,
0.007810309529304504,
0.0006952978437766433,
0.0035898799542337656,
0.022998588159680367,
0.007346570957452059,
0.027573615312576294,
0.02499254047870636,
-0.03303643316030502,
-0.026501864194869995,
0.... | 0.095955 |
the last attempt. ## Official provider plugins Official provider plugins publish their own model catalog rows. These providers require \*\*no\*\* `models.providers` model entries; enable the provider plugin, set auth, and pick a model. Use `models.providers` only for explicit custom providers or narrow request settings... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/model-providers.md | main | opebclaw | [
-0.04630181938409805,
-0.04715676233172417,
-0.07546466588973999,
0.054896533489227295,
0.03886927291750908,
-0.05021686479449272,
-0.026447931304574013,
-0.02386736497282982,
-0.014415214769542217,
-0.003051570849493146,
0.06868352741003036,
-0.04026777297258377,
0.023282794281840324,
0.0... | 0.011378 |
- Policy note: OpenAI Codex OAuth is explicitly supported for external tools/workflows like OpenClaw. - For the common subscription plus native Codex runtime route, sign in with `openai` auth and configure `openai/gpt-5.5`; OpenAI agent turns select Codex by default. - Use provider/model `agentRuntime.id: "openclaw"` o... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/model-providers.md | main | opebclaw | [
-0.08943237364292145,
0.0008300690678879619,
-0.07831455022096634,
0.023230068385601044,
0.05262797325849533,
-0.08063647150993347,
-0.05724110081791878,
0.010416862554848194,
-0.0215658787637949,
-0.03481303155422211,
0.003800459671765566,
0.02308153174817562,
-0.027741990983486176,
0.077... | 0.035182 |
- Auth: `KILOCODE\_API\_KEY` - Example model: `kilocode/kilo/auto` - CLI: `openclaw onboard --auth-choice kilocode-api-key` - Base URL: `https://api.kilo.ai/api/gateway/` - Static fallback catalog ships `kilocode/kilo/auto`; live `https://api.kilo.ai/api/gateway/models` discovery can expand the runtime catalog further.... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/model-providers.md | main | opebclaw | [
-0.05979166552424431,
-0.029341842979192734,
-0.08194874227046967,
0.07858891785144806,
-0.037171151489019394,
-0.06575214117765427,
-0.05473246052861214,
-0.029698006808757782,
-0.012163398787379265,
-0.031925734132528305,
0.03375338762998581,
-0.06560484319925308,
0.009899522177875042,
0... | 0.046629 |
`grok-3-mini`, `grok-4`, and `grok-4-0709` to their `\*-fast` variants. `tool\_stream` defaults on; disable via `agents.defaults.models["xai/"].params.tool\_stream=false`. Ships as the bundled `cerebras` provider plugin. GLM uses `zai-glm-4.7`; OpenAI-compatible base URL is `https://api.cerebras.ai/v1`. ## Providers vi... | https://github.com/openclaw/openclaw/blob/main/docs/concepts/model-providers.md | main | opebclaw | [
-0.06543845683336258,
-0.04043227434158325,
-0.0809396281838417,
-0.021375712007284164,
0.005995444022119045,
-0.06955611705780029,
-0.05095360055565834,
0.014606747776269913,
0.0465201772749424,
-0.03326811268925667,
0.03596527501940727,
-0.08832135796546936,
-0.014445694163441658,
0.0511... | 0.122442 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.