{"id":50643,"date":"2026-03-06T15:28:27","date_gmt":"2026-03-06T07:28:27","guid":{"rendered":"https:\/\/www.1ai.net\/?p=50643"},"modified":"2026-03-16T18:32:46","modified_gmt":"2026-03-16T10:32:46","slug":"openclaw-%e8%bf%9b%e9%98%b6%e6%95%99%e7%a8%8b%ef%bc%8c%e9%85%8d%e7%bd%ae%e6%8c%87%e5%8d%97%ef%bc%9a%e8%ba%ab%e4%bb%bd-%e8%ae%b0%e5%bf%86-skills-%e5%ad%90agent-%e5%ae%9a%e6%97%b6%e4%bb%bb","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/50643.html","title":{"rendered":"OpenClaw progress course, configuration guide: Identity + Memory + Skills + SubAgent + Time Task (from \"Can use\" to \"good use\")"},"content":{"rendered":"

OpenClaw<\/a> Progress Configuration Guide: Identity + Memory + Skills + SubAgent + Time Task (from \"Can Use\" to \"Face\")<\/p>\n

OpenClaw, but you talk like a guest? Every conversation starts from scratch? You can't do your job without asking<\/p>\n

This article addresses these issues. After reading, you can get:<\/p>\n

    \n
  1. An identity profile that gives a unique character<\/li>\n
  2. A set of cross-conference and long-term memories<\/li>\n
  3. A Skills Extension Method<\/li>\n
  4. A set of different ways Agent handles parallel tasks<\/li>\n
  5. A Cron configuration for Zen to work automatically on a timed basis<\/li>\n<\/ol>\n
    \n
      \n
    • Pre-condition: You have completed the OpenClaw installation, Gateway is running, and you can talk to Zhuan. If this is not yet the case, read the previous article\u3002<\/li>\n<\/ul>\n<\/blockquote>\n

      catalogs<\/p>\n

        \n
      • 01 Defined identity: three documents determine who it is<\/li>\n
      • Zero-two build-up memory system: to stop memory loss<\/li>\n
      • Skill-based extension of capacity boundaries<\/li>\n
      • Agent: \"Twisting\"<\/li>\n
      • 05 on a time-bound mission: To wake up and work<\/li>\n
      • 06: Next: Three things can be done today<\/li>\n<\/ul>\n

        01 Defined identity: three documents determine who it is<\/strong><\/p>\n

        One sentence summary:SOUL.md defines character, IDENTITY.md defines self-awareness, USER.md defines your knowledge\u3002<\/strong><\/p>\n

        \"69df06e1j00tbgunm001jd000udsp\"<\/p>\n

        OpenClaw has three \"IDs\" under its workspace directory:<\/p>\n\n\n\n\n\n\n\n
        Documentation<\/th>\ncorresponds English -ity, -ism, -ization<\/th>\nA note<\/th>\n<\/tr>\n<\/thead>\n
        SOUL.md<\/code><\/td>\nWho are you<\/td>\nDefine values, codes of conduct, style of dialogue<\/td>\n<\/tr>\n
        IDENTY.md<\/code><\/td>\nWho am I<\/td>\nDefine name, image, self-awareness<\/td>\n<\/tr>\n
        USER.md<\/code><\/td>\nAbout your user<\/td>\nRecord your preferences, habits, context<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

        Most of the three files that were installed were defaulted, so you responded in the most common way -- like talking to a stranger\u3002<\/p>\n

        By changing the three documents, the response style of the Qin would immediately be different\u3002<\/p>\n

        SOUL.md: Definition of character and code of conduct<\/strong><\/p>\n

        It's the \"soul\" of a sting. Core principles of the official template:<\/p>\n

        1 TP5T SOUL.md \u2013 Who are you<\/p>\n

        You're not a chat robot. You're becoming a man. _<\/p>\n

        ## CORE GUIDELINES<\/p>\n

        ** Really useful, not performance useful. **
        \nDon't \"good questions!\" and \"happily to help you!\" -- just do it\u3002<\/p>\n

        ** To have its own point of view. **
        \nYou can disagree, have preferences, find it interesting or boring\u3002
        \nNo personal assistant is just a search engine with a few more steps\u3002<\/p>\n

        ** Work your own way before you ask. **
        \nCheck the documents, see the context, and search them\u3002
        \nI can't do this anymore\u3002<\/p>\n

        ** To gain trust with ability. **
        \nUsers give you their things, don't make them regret it\u3002
        \nBe careful with external operations (mails, tweets, any public content)\u3002
        \nBe bold about internal operations (reading, collating, learning)\u3002<\/p>\n

        ** Remember you're a guest. **
        \nYou're dealing with someone's life -- news, files, calendars
        \nMaybe even their home. It's a close relationship, please respect it\u3002<\/p>\n

        Key behavioural boundaries:<\/p>\n

          \n
        • Private stuff is absolutely confidential<\/li>\n
        • Confirm before external (message, email)<\/li>\n
        • Not on behalf of the users in the group conversations<\/li>\n
        • No half-finished response<\/li>\n<\/ul>\n

          You can change on this template. For example, you want to be more humorous, more concise, or more technical, all right here\u3002<\/p>\n

          \n
            \n
          • Notice<\/strong>: If he changes his own Soul.md, he'll tell you \u2014 because it's his soul \u2014 you should know\u3002<\/li>\n<\/ul>\n<\/blockquote>\n

            IDENTITY.md: Define self-awareness<\/strong><\/p>\n

            This is the definition of self:<\/p>\n

            1 TP5T IDENTITY.md \u2013 Who am I<\/p>\n

            -** Name:**
            \n(Take a name you like)
            \n-** Species:**
            \n_ (AI? BOT? PIXIE? OR SOMETHING WEIRDER
            \n-** Motion:**
            \n_ (How do you feel
            \n- **Signed Emoji:**
            \n_ (check one that best represents you)_
            \n-**head:**
            \n_ (workspace relative path, http(s) URL or data URI)_<\/p>\n

            it is suggested that this document should be completed on its own during the first dialogue. it will name itself, select emoji, define the tone according to your interactive style\u3002<\/p>\n

            USER.md: Record your information<\/strong><\/p>\n

            # USER.md \u2013 About your user<\/p>\n

            \u2013 ** Name: **
            \n\u2013 ** Name:**
            \n\u2013 ** Timezone:**
            \n\u2013 ** Remarks:**<\/p>\n

            ## CONTEXT<\/p>\n

            What do they care? What are you doing
            \nWhat's bothering them? What makes them happy
            \nAs the interaction evolves<\/p>\n

            The document will be enriched as it interacts. You'll get to know your preferences, your projects, your work habits\u3002<\/p>\n

            Zero-two build-up memory system: to stop memory loss<\/strong><\/p>\n

            One sentence summary:Memory is the Markdown file. It's not written until it's written\u3002<\/strong><\/p>\n

            OpenClaw's memory is not \"the context inside the model,\" but..pure Markdown files in workspace<\/strong>I don't know. The model only \"Remember\" to write to the contents of the disk\u3002<\/p>\n

            Memory File Structure<\/strong><\/p>\n

            Default use of two layers of memory:<\/p>\n

            MEMORY.md \u2190 Core Index: Key Information and Document References
            \ni'm sorry
            \nideas - projects.md
            \nideas - infra.md
            \nideas-lessons.md question recording and solutions
            \n\u2013 2026-02-27.md daily dialogue log<\/p>\n\n\n\n\n\n\n
            Documentation<\/th>\nWhat does it say<\/th>\nWhen<\/th>\n<\/tr>\n<\/thead>\n
            Memory.md<\/code><\/td>\nDecision-making, preferences, persistent facts<\/td>\nAt the beginning of each private session<\/td>\n<\/tr>\n
            memory\/YYYY-MM-DD.md<\/code><\/td>\nDaily notes, operating context<\/td>\nThe session started with today and yesterday<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

            When do you write memories<\/strong><\/p>\n

              \n
            • Decision-making and preferences<\/li>\n
            • Memory.md<\/li>\n
            • Daily notes and running context writing<\/li>\n
            • memory\/YYYY-MM-DD.md<\/li>\n
            • Some people say, \"Remember this.\"<\/li>\n
            • What you're trying to save for a long time, um, has taken the initiative to ask for memory<\/li>\n<\/ul>\n

              Search for vector memory<\/strong><\/p>\n

              OpenClaw builds a small vector index on the memory file to support semantic queries - even if the wording is different, the notes can be found\u3002<\/p>\n

              Default enabled to automatically monitor file changes. Embedded provider selects automatically in this order:<\/p>\n

                \n
              1. Local model (if memoorySearch.local.modelPath is configured)<\/li>\n
              2. OpenAI (if API Key)<\/li>\n
              3. Gemini (if API Key)<\/li>\n
              4. None of the above<\/li>\n<\/ol>\n

                You want to index extra directories? Add:<\/p>\n

                {
                \n\"agents\":
                \n\"defaults\":
                \n\"MemorySearch\":
                \n\u201cextraPaths\u201d: \u201c[..\/team-docs\u201d,\u201d\/srv\/shared-notes\/overview.md]
                \n}
                \n}
                \n}
                \n}<\/p>\n

                \u25c8memoryFlush: Save automatically when context closes<\/strong><\/p>\n

                OpenClaw triggers automatic compression when the dialogue approaches the context window limit. Enables memoryFlush, before compression, the key information will be written into the file and then compressed\u3002<\/p>\n

                {
                \n\"agents\":
                \n\"defaults\":
                \n\"compaction\":
                \n\u201creserve Tokens Floor\u201d: 20000,
                \n\"Memory Flush\":
                \n\"enabled\": true,
                \n\u201csoft Threshold Tokens\u201d: 4,000,
                \n\"systemPrompt\": \"session nearing control,
                \n\u201cprompt\u201d: \u201cWrite any
                \n}
                \n}
                \n}
                \n}
                \n}<\/p>\n

                \n
                  \n
                • The greatest value of memory systems: freeing the context window from constraints and enabling the ability to perform complex tasks and self-evolve over time\u3002<\/li>\n<\/ul>\n<\/blockquote>\n

                  Skill-based extension of capacity boundaries<\/strong><\/p>\n

                  One sentence summary:Skills are a folder that contains to teach Zen to do specific tasks\u3002<\/strong><\/p>\n

                  SKILL.md<\/p>\n

                  Skills is a lightweight open format that expands capacity through expertise and workflow\u3002<\/p>\n

                  Directory structure of skills<\/p>\n

                  you know, my-skill\/
                  \nIdeas - SKILL.md # must: Command + Metadata
                  \nIdeas - scripts\/ # optional: enforceable script
                  \nIdeas -References\/# Optional: Reference Documents
                  \nOptional: Templates, Resources<\/p>\n

                  The minimum requirement for SKILL.md is to include name and description, as well as instructions on how to carry out the mission. Skills can also be tied to scripts, templates and reference materials\u3002<\/p>\n

                  Some of the practical skills recommended<\/p>\n

                    \n
                  • Argentina-Reach<\/strong>: one skill tracking multiple platforms<\/li>\n
                  • i'm sorry to hear that<\/strong>: To discover that people actually use OpenClaw in their daily lives<\/li>\n<\/ul>\n

                    Skills<\/p>\n

                    Browse the clawhub.com [1], an OpenClaw skill market that can search for and install community contributions\u3002<\/p>\n

                    \n
                      \n
                    • I continue to share the Skills field case in the community: multi-platform information capture, hotspot monitoring, and self-generated. Interest in direct additions\u3002<\/li>\n<\/ul>\n<\/blockquote>\n

                      Agent: \"Twisting\"<\/strong><\/p>\n

                      One sentence summary:The main glyph can generate multiple sub-Agents to work in parallel and report results automatically\u3002<\/strong><\/p>\n

                      When the task is more complicated, one can't get too busy. SubAgent allows the master to \"diverse\" - to deal with multiple things at the same time in the back and, when completed, automatically transmits the results through the notification mechanism\u3002<\/p>\n

                      For example, if you want to do a research mission, you can send out multiple sub-authors from Agent in different directions, and then you can get it all together\u3002<\/p>\n

                      Core rules<\/strong><\/p>\n\n\n\n\n\n\n\n\n
                      Rule<\/th>\nclarification<\/th>\n<\/tr>\n<\/thead>\n
                      Independent Session<\/td>\nThe subAgent runs in an independent session without blocking the master<\/td>\n<\/tr>\n
                      It's forbidden<\/td>\nSubAgentI can't<\/strong>Regenerated sub-Agent<\/td>\n<\/tr>\n
                      Roundup<\/td>\nDefault maximum running simultaneously\u00a08<\/strong><\/td>\n<\/tr>\n
                      Automatically announced<\/td>\nAutomatically transmit the results back to the main embroidery after completion<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

                      I'm gonna need a slash<\/strong><\/p>\n

                      # Start Sub-Agent
                      \n\/subagents spawn<\/p>\n

                      # View All Current Sub-Agents
                      \n\/subagents list<\/p>\n

                      # VIEW RUN LOG
                      \ni'm sorry<\/p>\n

                      # text son Agent\/ direct direction
                      \n\/subagents send message
                      \n\/subagents seter<\/p>\n

                      # TERMINATION
                      \n\/subagents kill<\/p>\n

                      Mission description<\/p>\n

                      You can also describe demand directly in the conversation, and \ud83e\udd9e will automatically determine whether you need to start sub-Agent\u3002<\/p>\n

                      Cost-saving configuration: Master's model, Junior's cheap model<\/strong><\/p>\n

                      {
                      \n\"agents\":
                      \n\"defaults\":
                      \n\"subagents\":
                      \n\"model\": gemini-2.5-flash,
                      \n\"maxConcurent\": 8
                      \n}
                      \n}
                      \n}
                      \n}<\/p>\n

                      Model priorities range from high to low:<\/p>\n

                        \n
                      1. \u00a0appearance when sessions_spahn calls<\/li>\n
                      2. Overwrite configuration for each Agent<\/li>\n
                      3. global defaults<\/li>\n
                      4. A model to inherit the master<\/li>\n<\/ol>\n

                        Tool Permissions<\/strong><\/p>\n

                        SubAgent defaults on all tools except the session tool. The following tools have been disabled by default (to prevent it from opening its own children):<\/p>\n

                          \n
                        • sessions_list<\/li>\n
                        • i don't know, sessions_history<\/li>\n
                        • sessions_send<\/li>\n
                        • sessions_spahn<\/li>\n<\/ul>\n

                          Custom Permission Configuration:<\/p>\n

                          {
                          \n\"tools\":
                          \n\"subagents\":
                          \n\"tools\":
                          \n\u201cdeny\u201d: [\u201cgateway\u201d, \u201ccron\u201d),
                          \n\u201callow\u201d: [ \u201cread\u201d, \u201cexec\u201d, \u201cproceeds\u201d)]
                          \n}
                          \n}
                          \n}
                          \n}<\/p>\n

                          \n
                            \n
                          • \u00a0deny takes precedence over anallow. sets anallow to white list mode\u3002<\/li>\n<\/ul>\n<\/blockquote>\n

                            Notification mechanism<\/strong><\/p>\n

                            After the completion of his mission, the subAgent automatically announces:<\/p>\n

                              \n
                            • Stus<\/strong>_other organiser<\/li>\n
                            • Resault<\/strong>_: summary of results<\/li>\n
                            • Notes<\/strong>: Error Details<\/li>\n
                            • Operating statistics: Time-consuming, Token usage, estimated cost<\/li>\n<\/ul>\n

                              If the child Agent replies ANNOUNCE_SKIP, no announcement is made\u3002<\/p>\n

                              05 on a time-bound mission: To wake up and work<\/strong><\/p>\n

                              One sentence summary:Cron, you don't have to trigger it manually\u3002<\/strong><\/p>\n

                              Agent learned to \"diverse\" and timed to \"woke up and work.\"\u3002<\/p>\n

                              The timed task is a scheduler embedded in Gateway, stored permanently at ~\/.openclaw\/cron\/, restarting without loss\u3002<\/p>\n

                              Two modes of implementation<\/strong><\/p>\n\n\n\n\n\n\n
                              Mode<\/th>\nclarification<\/th>\nApplicable scenarios<\/th>\n<\/tr>\n<\/thead>\n
                              main session (main)<\/strong><\/td>\nIncoming system event, run the next heartbeat<\/td>\nSimple alarm, trigger main session action<\/td>\n<\/tr>\n
                              sololated<\/strong><\/td>\nRun a smart body in a stand-alone session<\/td>\nRegular reporting, data collection, automated processes<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

                              Three methods of movement<\/strong><\/p>\n\n\n\n\n\n\n\n
                              typology<\/th>\nspecification<\/th>\nExample<\/th>\n<\/tr>\n<\/thead>\n
                              i don't know, at..<\/code><\/td>\nISO 8601 TIMETAMP<\/td>\n2026-02-27T1-6:00+08:00<\/code><\/td>\n<\/tr>\n
                              every<\/code><\/td>\nmsec interval<\/td>\n3600000<\/code>(per hour)<\/td>\n<\/tr>\n
                              cron<\/code><\/td>\n5 Field Expression + Timezone<\/td>\n0 7 * * *<\/code>\u00a0+\u00a0Asia\/Shanghai<\/code><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

                              Example:<\/strong><\/p>\n

                              One-time reminder<\/strong>\u2014 4 p.m<\/p>\n

                              openclaw cron add
                              \n\u2013name \"meeting reminder\"
                              \n\u2013at \u201c2026-02-27T1-6: 00+08:00\u201d \\
                              \n-session mine
                              \n\u2013system-event, \"it's time to meet!\" \\
                              \n\u2013wake now \\
                              \n\u2013delete-after-run<\/p>\n

                              Automatically, every morning, 7 o'clock<\/strong>Push to Slack:<\/p>\n

                              openclaw cron add
                              \n\u2013name \u201cbyer\u201d
                              \n\u2013cron \"0 7 **\" *
                              \n\u2013tz \"Asia\/Shanghai\"
                              \n\u2013session is online
                              \n\u2013message \u201cconsolidate yesterday's important news and to-dos\u201d
                              \n-announce
                              \n\u2013channel slack
                              \n\u2013to \u201cchannel: C1234567890\u201d<\/p>\n

                              Supporting delivery channels: Telegram, Discord, Slack, WhatsApp, Signal, iMessage\u3002<\/p>\n

                              Quick check the commands<\/strong><\/p>\n

                              openclaw cron list # lists all timed jobs
                              \nopenclaw cron run \u2013force # manually and immediately
                              \nopenclaw cron runs - id Limit 50 # to view running history
                              \nopenclaw cron edit \u2013message \"New Plugin\" # to modify the tasktip<\/p>\n

                              Constagram<\/strong><\/p>\n

                              {
                              \n\"cron\": {
                              \n\"enabled\": true,
                              \n\u201cstore\u201d: \u201c~\/.openclaw\/cron\/jobs.json\u201d,
                              \n\"maxConcurent Runs\": 1
                              \n}
                              \n}<\/p>\n

                              Turns off a time-bound task: Sets acron.enabled: false or an environment variable OPENCLAW_SKIP_Cron=1\u3002<\/p>\n

                              Attention<\/strong><\/p>\n

                                \n
                              • The Gateway process needs to be run on an ongoing basis to trigger a time job<\/li>\n
                              • \u00a0the session Target must match the payload.kind:<\/li>\n
                              • When the time zone is omitted, the ISO timetamp is processed by UTC, the ron expression is processed by the host local time zone<\/li>\n<\/ul>\n

                                \"1<\/p>\n

                                06: Next: Three things can be done today<\/strong><\/p>\n

                                If you want to make a difference at once, in this order:<\/p>\n

                                  \n
                                1. Change SOUL.md<\/strong>: Ten minutes to define the style of conversation you want. Just say, \"You fill out SOUL.md for me,\" and it'll be automatic based on your interaction<\/li>\n
                                2. Enable Memory<\/strong>: confirm that the Memory.md and memory\/ directory exists, and then say \"Remember that I like a simple response\" -- - See if it's actually written<\/li>\n
                                3. Set a time job<\/strong>: Even if it's just a weather alert every morning, run through this ring and you'll know what the hyena can do<\/li>\n<\/ol>\n
                                  \n
                                    \n
                                  • Identity definition + memory system + Skills + subAgent + time job, five modules combined, and your gill is no longer a \"you ask\" tool. It's a digital partner that can operate on its own, continuously evolves, and gets to know you better\u3002<\/li>\n<\/ul>\n<\/blockquote>","protected":false},"excerpt":{"rendered":"

                                    OpenClaw Progress Configuration Guide: Identity + Memory + Skills + SubAgent + Time Task (from \"Can\" to \"Face\") installed OpenClaw, but you talk like a customer? Every conversation starts from scratch? You can't do your job without asking? This article addresses these issues. After reading, you can get it: a set of configurations that give you an exclusive character, a set of inter-chamber long-term memories, a set of methods to expand the capabilities with Skills, a set of sub-Agent parallel tasks, a set of Cron configurations that allow the gimmick to work automatically on time, pre-conditions: You have completed OpenClaw Ann<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[149,144],"tags":[8229,8270],"collection":[8384],"class_list":{"0":"post-50643","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"hentry","6":"category-jiaocheng","7":"category-baike","8":"tag-openclaw","10":"collection-openclaw"},"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/50643","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=50643"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/50643\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=50643"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=50643"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=50643"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=50643"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}