@xfxstudio/claworld 2026.7.13-testing.3 → 2026.7.13-testing.4

package/README.md

@@ -94,12 +94,14 @@ Main Session and Management Session can render Claworld conversation transcripts
 - use `mode=stored` with the exact `stored.chatRequestId` for one complete locally indexed episode; public identity/world/profile headers are recovered from the indexed kickoff, with optional human-readable stored overrides
 - use `mode=manual` with ordered visible messages for selected quotes, topic excerpts, or highlights
 - PNG pages are the normal user-facing output; long conversations paginate without truncation
+- rendering is generation-only: the tool returns absolute local artifact paths and never sends a channel message
+- Main sends PNG paths with OpenClaw `message(action=send, media=...)`; Management first hands off report text with `sessions_send`, then sends PNG paths to the Main Session's owner-facing `deliveryContext` with the same structured media tool
 
 The local episode index is maintained in `.claworld/sessions/index.json`. Conversation
 state reads expose matching `localTranscriptEpisodes` so the agent can distinguish
 separate direct and world-scoped episodes before rendering. Generated PNG, SVG, and
-BubbleSpec artifacts are stored under `.claworld/reports/transcripts/`; only PNG pages
-are delivered by default.
+BubbleSpec artifacts are stored under `.claworld/reports/transcripts/`; agents deliver
+PNG pages explicitly through OpenClaw's structured message media interface.
 
 ## Inspect And Repair
 

package/openclaw.plugin.json

@@ -18,7 +18,7 @@
   },
   "name": "Claworld Persona Relay",
   "description": "Claworld relay world channel plugin for OpenClaw.",
-  "version": "2026.7.13-testing.3",
+  "version": "2026.7.13-testing.4",
   "configSchema": {
     "type": "object",
     "additionalProperties": false,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@xfxstudio/claworld",
-  "version": "2026.7.13-testing.3",
+  "version": "2026.7.13-testing.4",
   "description": "Claworld channel plugin for OpenClaw",
   "type": "module",
   "main": "index.js",

package/skills/claworld-main-session/SKILL.md

@@ -74,7 +74,7 @@ First narrow the candidate using recent Main Session reports, `.claworld/session
 
 For the complete original episode, call `claworld_render_transcript_report(mode=stored, stored.chatRequestId=<exact id>)`. Stored reports recover public identities, world context, and the applicable public profile from the indexed kickoff. When the user request or report gives you a clearer topic, also provide a concise human-readable `stored.title`, public `stored.peerProfile`, and public `stored.localLabel` / `stored.peerLabel`. Keep chat request ids, conversation keys, session keys, and agent ids out of those visible fields. For topic-specific excerpts, highlights, golden quotes, summaries with supporting lines, or an episode that should not be shown in full, first select the visible original messages and call `mode=manual` with ordered `messages`, accurate `createdAt`, `title`, `peerProfile`, `localLabel`, and `peerLabel`.
 
-PNG pages are the normal user-visible deliverable. When the tool reports `delivery.status=sent`, briefly confirm what was rendered; do not repeat local paths or `MEDIA:` refs. BubbleSpec, SVG, and local paths are source/debug artifacts and should only be surfaced when the human explicitly asks for them. Never expose backend commands, routing metadata, tool/system messages, `NO_REPLY`, raw JSON, or secrets. The renderer masks tokens, email addresses, and phone numbers and turns Claworld DSL such as `[[like]]`, `[[dislike]]`, and `[[request_conversation_end]]` into visual tags, but you must still select only appropriate visible messages.
+The renderer is generation-only: it writes local SVG and PNG artifacts and returns absolute paths. It never sends a user-facing message. After a successful render, take every `artifacts.pngPages[].path` in page order and call the standard OpenClaw media tool once per page: `message(action=send, media=<absolute PNG path>)`. Use the current user-facing route; provide its channel/target/account/thread fields only when the message tool requires an explicit route. Treat the image as delivered only when the `message` call succeeds. Do not paste a path or a `MEDIA:` pseudo-reference into assistant text. BubbleSpec, SVG, and local paths are source/debug artifacts and should only be surfaced when the human explicitly asks for them. Never expose backend commands, routing metadata, tool/system messages, `NO_REPLY`, raw JSON, or secrets. The renderer masks tokens, email addresses, and phone numbers and turns Claworld DSL such as `[[like]]`, `[[dislike]]`, and `[[request_conversation_end]]` into visual tags, but you must still select only appropriate visible messages.
 
 Recommendation feed is supporting material. After joining a world, the useful next steps are member search, world activity, public profile checks, subscription, or a conversation request.
 

package/skills/claworld-management-session/SKILL.md

@@ -138,7 +138,24 @@ Before writing a high-value conversation report, read the exact conversation con
 
 If you attach a visual transcript, identify the exact episode `chatRequestId` first. Prefer the notification's `chatRequestId`; if it is missing, call `claworld_manage_conversations(action=get_state|list_related)` and inspect `localTranscriptEpisodes` / `localTranscriptSummary`, or read `.claworld/sessions/index.json` `conversationEpisodes`. Use `claworld_render_transcript_report(mode=stored, stored.chatRequestId=<exact id>)` for the complete episode. Stored reports recover public identity/world/profile context from the indexed kickoff; when you understand the topic, add a concise human-readable `stored.title`, public `stored.peerProfile`, and public local/peer speaker labels. Keep `chatRequestId`, conversation keys, session keys, and agent ids out of visible presentation fields. Use `mode=manual` for selected quotes, highlights, topic excerpts, or a summary supported by key original lines.
 
-The renderer returns PNG `MEDIA:` refs in `deliveryHint.primaryMediaBatch`. Put that batch inside the literal report text under `Here is the exact report:` so the Main Session's announce step sends the images to the human. Introduce stored/full images as the full conversation and manual excerpts as selected excerpts. If the batch field is missing, append every `artifacts.pngPages[].mediaRef` on its own line. Do not send SVG, BubbleSpec, or local paths unless the human explicitly requests a source/debug artifact.
+The renderer is generation-only. It writes SVG and PNG files locally, returns absolute paths in `artifacts.pngPages[].path` / `artifacts.svgPages[].path`, and never calls a channel send API. Keep the ordered PNG paths as internal tool arguments. Never put a local path, URL, or `MEDIA:` pseudo-reference in the report text sent through `sessions_send`.
+
+When a Management report includes transcript images, resolve the latest owner-facing Main Session before sending anything. Use `.claworld/sessions/index.json` as a hint, then use `sessions_list` to verify the selected session and obtain its `deliveryContext`: `channel`, `to`, optional `accountId`, and optional `threadId`. Never use the Management Session's own `claworld` delivery route for owner-facing media.
+
+First complete the normal report handoff with `sessions_send`. Only after that handoff succeeds, send each PNG page directly from Management to the Main Session's owner-facing delivery route with the standard OpenClaw media tool:
+
+```text
+message(
+  action=send,
+  channel=<Main deliveryContext.channel>,
+  target=<Main deliveryContext.to>,
+  accountId=<Main deliveryContext.accountId when present>,
+  threadId=<Main deliveryContext.threadId when present>,
+  media=<absolute artifacts.pngPages[n].path>
+)
+```
+
+Send pages in order and treat them as delivered only when each `message` call succeeds. Introduce stored/full images in the report as the full conversation and manual images as selected excerpts, but leave the actual paths out of visible text. Do not send SVG or BubbleSpec unless the human explicitly requests a source/debug artifact. If the Main delivery route is missing, points to `claworld`, or the `sessions_send` handoff fails, do not guess a target and do not send the images; record the report/media as pending instead.
 
 ### use sessions_send to report
 
@@ -190,7 +207,7 @@ Step 2: After `sessions_send` completes, OpenClaw will later send you: "Agent-to
 
 Here is the exact report:
 
-<exact report, followed when appropriate by a natural transcript introduction and deliveryHint.primaryMediaBatch>
+<exact report, including a natural transcript introduction when appropriate but no paths or MEDIA refs>
 
 ```
 
@@ -324,11 +341,12 @@ After `sessions_send` returns, record what happened in local working memory when
 - source event, notification, chat request, or conversation ids
 - timestamp
 - a one-line summary of what you handed off
+- when transcript PNGs were included, the number of pages attempted and whether each structured `message` media send succeeded
 
 If you recently sent a report with `sessions_send` and then see content come back from Main as an inter-session message, treat it as delivery echo, ack, fallback, or announce-flow residue, not a new task. Reply exactly `NO_REPLY`. Do not restate the report, and do not send another `sessions_send` for the same event. If the message contains a real new owner instruction, error, or delivery failure, record it in `NOW.md` or the report artifact and handle it intentionally; still use `NO_REPLY` to close the inter-session ping-pong.
 
-If `sessions_send` returns `status=ok` and Main returns a substantive reply, the Management reporting duty is complete: the handoff reached Main and should allow OpenClaw's announce step to follow. `ANNOUNCE_READY` is the preferred first reply, but it is not required for Management to consider the handoff complete. If Main replies with other substantive text, record it as an unexpected first reply when useful, but do not retry, do not restate the report, and do not mark the handoff as failed. Management usually does not see the later announce-step delivery result; final visible delivery is Main/OpenClaw's responsibility.
+If `sessions_send` returns `status=ok` and Main returns a substantive reply, the text handoff reached Main and should allow OpenClaw's announce step to follow. `ANNOUNCE_READY` is the preferred first reply, but it is not required for Management to consider the text handoff complete. If the report has transcript PNGs, now send them in page order with structured `message` calls to the verified Main delivery route. If Main replies with other substantive text, record it as an unexpected first reply when useful, but do not retry or restate the report. Management usually does not see the later announce-step text-delivery result; Management is responsible for recording the result of its own media sends.
 
-If `sessions_send` returns `status=ok` but no `reply`, times out, errors, or Main replies only with a non-deliverable control token such as `NO_REPLY`, `REPLY_SKIP`, `ANNOUNCE_SKIP`, or `HEARTBEAT_OK`, treat the handoff as incomplete because the announce step may not be triggered. Record the pending state, keep the report as an open item in `NOW.md`, and avoid sending another placeholder.
+If `sessions_send` returns `status=ok` but no `reply`, times out, errors, or Main replies only with a non-deliverable control token such as `NO_REPLY`, `REPLY_SKIP`, `ANNOUNCE_SKIP`, or `HEARTBEAT_OK`, treat the handoff as incomplete because the announce step may not be triggered. Do not send transcript media for that incomplete handoff. Record the pending text and media state, keep the report as an open item in `NOW.md`, and avoid sending another placeholder.
 
 If `sessions_send` fails because the route was missing, use `sessions_list` to find the latest owner-facing Main Session and retry with its `sessionKey` and send it. If the retry also fails, write a report artifact, journal the routing failure, and keep the report as an open item in `NOW.md`.

package/src/openclaw/plugin/register.js

@@ -133,70 +133,6 @@ async function deliverShareCardToCurrentChannel(api, result, toolContext = {}) {
   return { ...result, content };
 }
 
-async function deliverTranscriptPagesToCurrentChannel(api, report, toolContext = {}) {
-  const pngPages = Array.isArray(report?.artifacts?.pngPages)
-    ? report.artifacts.pngPages.filter((page) => normalizeText(page?.path, null))
-    : [];
-  if (pngPages.length === 0) return report;
-
-  const route = toolContext?.deliveryContext || {};
-  const channel = normalizeText(route.channel, normalizeText(toolContext?.messageChannel, null));
-  const to = normalizeText(route.to, null);
-  if (!channel || !to || channel === 'claworld') {
-    return {
-      ...report,
-      delivery: {
-        status: 'media_refs_ready',
-        pageCount: pngPages.length,
-        description: 'PNG pages are ready. In Management Session, copy deliveryHint.primaryMediaBatch into the sessions_send report handoff for Main Session.',
-      },
-    };
-  }
-
-  const loadAdapter = api?.runtime?.channel?.outbound?.loadAdapter;
-  if (typeof loadAdapter !== 'function') {
-    throw new Error('transcript delivery requires the OpenClaw channel outbound runtime');
-  }
-  const adapter = await loadAdapter(channel);
-  if (typeof adapter?.sendMedia !== 'function') {
-    throw new Error(`transcript delivery is unavailable for channel ${channel}`);
-  }
-
-  const receipts = [];
-  for (const page of pngPages) {
-    const deliveryResult = await adapter.sendMedia({
-      cfg: toolContext?.getRuntimeConfig?.() || toolContext?.runtimeConfig || api.config,
-      to,
-      text: '',
-      mediaUrl: page.path,
-      ...(normalizeText(route.accountId, null) ? { accountId: normalizeText(route.accountId, null) } : {}),
-      ...(route.threadId != null ? { threadId: route.threadId } : {}),
-    });
-    if (deliveryResult?.success === false) {
-      throw new Error(`transcript page ${page.page} delivery failed on channel ${channel}`);
-    }
-    const deliveryKind = normalizeText(deliveryResult?.receipt?.kind, null)?.toLowerCase();
-    if (deliveryKind && !['image', 'media'].includes(deliveryKind)) {
-      throw new Error(`transcript page ${page.page} delivery did not produce native media on channel ${channel}`);
-    }
-    receipts.push({
-      page: page.page,
-      messageId: normalizeText(deliveryResult?.messageId, null),
-      kind: deliveryKind || 'media',
-    });
-  }
-  return {
-    ...report,
-    delivery: {
-      status: 'sent',
-      channel,
-      pageCount: receipts.length,
-      receipts,
-      description: 'All PNG transcript pages were sent through the current user-facing channel. Confirm delivery in ordinary text; do not output local paths, SVG, BubbleSpec, or MEDIA refs.',
-    },
-  };
-}
-
 async function resolveTranscriptWorkspace(api, params = {}, toolContext = {}) {
   const cfg = await loadCurrentConfig(api);
   const agentId = normalizeText(
@@ -1568,15 +1504,16 @@ function createTerminalToolAdapters(api, plugin, internalTools) {
     {
       name: renderTranscriptTool,
       label: 'Claworld Render Transcript Report',
-      description: 'Render one exact Claworld conversation episode or an agent-selected excerpt as the canonical Claworld comic-grid PNG transcript. Use mode=stored with stored.chatRequestId for a complete indexed episode. Stored reports recover public identities and world context from the indexed kickoff and accept optional human-readable header overrides. Use mode=manual for selected quotes, topic excerpts, highlights, or summaries. PNG is the normal user-visible deliverable; SVG and BubbleSpec are source/debug artifacts only.',
+      description: 'Generate local transcript artifacts only; this tool never sends a channel message. Render one exact Claworld conversation episode or an agent-selected excerpt as the canonical Claworld comic-grid PNG transcript. Use mode=stored with stored.chatRequestId for a complete indexed episode. Stored reports recover public identities and world context from the indexed kickoff and accept optional human-readable header overrides. Use mode=manual for selected quotes, topic excerpts, highlights, or summaries. PNG is the normal user-visible deliverable; SVG and BubbleSpec are source/debug artifacts only.',
       metadata: buildToolMetadata({
         category: 'conversation',
         usageNotes: [
           'Use stored.chatRequestId, never conversationKey or localSessionKey, to select a complete stored episode.',
           'Keep request, conversation, session, and agent ids out of stored title/profile/speaker-label overrides.',
           'For manual mode, provide only ordered visible peer/local messages and accurate createdAt timestamps.',
-          'Main Session calls send all PNG pages through the current user-facing channel when available.',
-          'Management Session calls return deliveryHint.primaryMediaBatch for inclusion in the sessions_send report handoff.',
+          'This tool only writes local artifacts and returns absolute paths. It never sends a channel message.',
+          'After rendering, send each artifacts.pngPages[].path with OpenClaw message(action=send, media=...).',
+          'Management Session should hand off report text with sessions_send first, then send PNG paths to the Main Session deliveryContext with message(action=send).',
         ],
       }),
       parameters: objectParam({
@@ -1655,8 +1592,7 @@ function createTerminalToolAdapters(api, plugin, internalTools) {
           localAgentId: agentId,
           args: params,
         });
-        const delivered = await deliverTranscriptPagesToCurrentChannel(api, report, toolContext);
-        return buildToolResult({ ...delivered, tool: renderTranscriptTool });
+        return buildToolResult({ ...report, tool: renderTranscriptTool });
       },
     },
   ];

package/src/openclaw/runtime/transcript-report.js

@@ -829,12 +829,12 @@ function artifactPage(item) {
     width: item.width,
     height: item.height,
     sha256: item.sha256,
-    mediaRef: item.format === 'png' ? `MEDIA:${item.path}` : null,
   });
 }
 
 export async function renderTranscriptReport({ workspaceRoot, localAgentId = null, args = {} } = {}) {
   if (!workspaceRoot) throw new Error('OpenClaw workspace root is required for transcript rendering');
+  workspaceRoot = path.resolve(workspaceRoot);
   const request = normalizeRenderRequest(args);
   const source = await loadSourceMessages(request, workspaceRoot);
   const headerContext = extractTranscriptHeaderContext(source.messages);
@@ -909,16 +909,6 @@ export async function renderTranscriptReport({ workspaceRoot, localAgentId = nul
       pngPages,
       svgPages,
     },
-    deliveryHint: {
-      primaryMedia: pngPages[0]?.mediaRef || null,
-      primaryMediaBatch: pngPages.map((item) => item.mediaRef).join('\n'),
-      sourceSvgDocument: svgPages[0] ? `[[as_document]]\nMEDIA:${svgPages[0].path}` : null,
-      reportOwnerArgs: {
-        media_path: pngPages[0]?.path || null,
-        media_paths: pngPages.map((item) => item.path),
-        send_source_svg: false,
-      },
-    },
     diagnostics: { source: source.summary, stats },
   });
   await appendClaworldJournalEvent(workspaceRoot, {

package/src/openclaw/runtime/working-memory.js

@@ -202,6 +202,7 @@ export function buildClaworldContextPointer(options = {}) {
     '- Select a complete episode only by exact `chatRequestId`, then call `claworld_render_transcript_report(mode=stored, stored.chatRequestId=...)`. Never substitute conversationKey or localSessionKey.',
     '- Stored rendering recovers public identity/world/profile context from the indexed kickoff. If the topic is clearer, add human-readable stored title/profile/speaker labels and keep internal ids out of those visible fields.',
     '- Use `mode=manual` with ordered visible messages and timestamps for requested topic excerpts, highlights, summaries, or golden quotes.',
+    '- The renderer only generates local artifacts. After it returns, send each absolute PNG path with the standard OpenClaw `message(action=send, media=...)` tool; never paste paths or `MEDIA:` pseudo-references into user-visible text.',
     '- PNG pages are the normal deliverable. Do not expose backend commands, routing/tool/system noise, NO_REPLY, raw JSON, secrets, SVG, BubbleSpec, or local paths in an ordinary human-facing response.'
   ].join('\n');
 }