docs: add TweetClaw workflow and repair Wiki safety

README.md

@@ -1,7 +1,5 @@
 <h1 align="center">Membase Plugin for OpenClaw</h1>
 
-[![Membase banner](https://github.com/user-attachments/assets/19393af8-7af0-4e8f-9967-b5c9d8119d83)](https://membase.so/?utm_source=github&utm_medium=openclaw-membase)
-
 <p align="center">
   Persistent long-term memory for OpenClaw using hybrid vector search and a knowledge graph.
 </p>
@@ -20,7 +18,7 @@
 
 Give your [OpenClaw](https://openclaw.ai/) agent persistent memory that survives across sessions. Membase uses hybrid vector search + knowledge graph to remember not just text, but entities, relationships, and facts.
 
-> **Free to start** — Sign up at [app.membase.so](https://app.membase.so) and connect in under a minute.
+> **Free to start** - Sign up at [app.membase.so](https://app.membase.so) and connect in under a minute.
 
 ## Install
 
@@ -77,7 +75,7 @@ The agent uses these tools autonomously during conversations:
 
 | Tool | Description |
 | --- | --- |
-| `membase_search` | Search memories by semantic similarity. Supports date filtering (`date_from`, `date_to`, `timezone`) and source filtering (`sources` — integrations, AI clients including `codex`/`hermes`, imports, `notion`, and other sources). Returns episode bundles with related facts and relevance scores. |
+| `membase_search` | Search memories by semantic similarity. Supports date filtering (`date_from`, `date_to`, `timezone`) and source filtering (`sources` - integrations, AI clients including `codex`/`hermes`, imports, `notion`, and other sources). Returns episode bundles with related facts and relevance scores. |
 | `membase_get_current_date` | Return runtime local and UTC time before converting relative dates into `date_from`/`date_to`. |
 | `membase_store` | Save important information to long-term memory. Proactively stores preferences, goals, and context. |
 | `membase_forget` | Delete a memory. Shows matches first, then deletes after user confirmation (two-step). |
@@ -87,10 +85,14 @@ The agent uses these tools autonomously during conversations:
 | `membase_update_wiki` | Update title/content/Project for an existing wiki document. Set Project to `null` to move to Basic. Reports the returned destination. |
 | `membase_delete_wiki` | Delete a wiki document with confirmation flow. |
 
+## Example Workflows
+
+- [Remember reviewed public X/Twitter signals from TweetClaw](./examples/tweetclaw-x-memory.md) - gather public X context with TweetClaw, then store reviewed summaries and source URLs in Membase Wiki.
+
 ## CLI Commands
 
 ```bash
-openclaw membase login              # OAuth login (PKCE) — opens browser
+openclaw membase login              # OAuth login (PKCE) - opens browser
 openclaw membase logout             # Remove stored tokens
 openclaw membase search <query>     # Search memories
 openclaw membase search <query> -s slack,gmail  # Filter by source
@@ -175,14 +177,14 @@ bun run build
 
 ## Contributing
 
-Contributions welcome — see [CONTRIBUTING.md](./CONTRIBUTING.md) for guidelines.
+Contributions welcome - see [CONTRIBUTING.md](./CONTRIBUTING.md) for guidelines.
 
 ## Links
 
-- [Membase](https://membase.so/?utm_source=github&utm_medium=openclaw-membase) — Website
-- [Dashboard](https://app.membase.so) — Manage your memories
-- [Docs](https://docs.membase.so) — Full documentation
-- [OpenClaw](https://openclaw.ai/) — AI agent framework
+- [Membase](https://membase.so/?utm_source=github&utm_medium=openclaw-membase) - Website
+- [Dashboard](https://app.membase.so) - Manage your memories
+- [Docs](https://docs.membase.so) - Full documentation
+- [OpenClaw](https://openclaw.ai/) - AI agent framework
 
 ## License
 

examples/tweetclaw-x-memory.md

@@ -0,0 +1,63 @@
+# Remember Reviewed X/Twitter Signals From TweetClaw
+
+Use this workflow when the same OpenClaw workspace needs structured public X/Twitter data and persistent memory. TweetClaw gathers the public source material, then Membase stores reviewed summaries, source URLs, and follow-up decisions.
+
+## Install The Plugins
+
+```bash
+openclaw plugins install clawhub:@xquik/tweetclaw
+openclaw plugins install @membase/openclaw-membase
+openclaw config set tools.alsoAllow '["explore", "tweetclaw", "openclaw-membase"]'
+```
+
+Keep credentials in the plugin that owns them:
+
+- TweetClaw keeps `XQUIK_API_KEY` in its OpenClaw plugin configuration. Never
+  paste the key into chats, prompts, logs, or memory.
+- Membase keeps OAuth tokens in its `tokenFile`, outside the plugin directory.
+- Do not store API keys, OAuth tokens, direct message contents, or raw private account exports in memory.
+
+## Capture Public Signals
+
+Use TweetClaw's free `explore` tool to identify the right endpoint, then use
+the optional `tweetclaw` tool for public X/Twitter source material:
+
+- Search tweets or search tweet replies for a launch, incident, competitor, keyword, or support topic.
+- Monitor tweets and webhooks for recurring public mentions.
+- Export public follower context or run user lookup when the audience or author matters.
+- Record giveaway draw results with the source tweet URL and draw criteria.
+- Use post tweets or post tweet replies only after the OpenClaw user approves the visible action.
+
+## Store A Reviewed Wiki Document
+
+After reviewing the TweetClaw result, use `membase_add_wiki` to store a compact
+document in a Project such as `X Research`. Include:
+
+- Source: `tweetclaw`
+- Capture date
+- Tweet URL or tweet ID
+- Author handle when relevant
+- Short summary
+- Sentiment or category if useful
+- Next action or decision
+
+Example memory content:
+
+```text
+Source: TweetClaw
+Captured: YYYY-MM-DD
+Topic: billing launch feedback
+Source URLs: https://x.com/example/status/123
+Summary: Users asked whether annual plans include API credits.
+Decision: Update the launch FAQ before the next announcement.
+```
+
+## Recall Later
+
+Use `membase_search_wiki` to recover the reviewed X/Twitter context without
+fetching every source again. `autoWikiRecall` can inject relevant Wiki context
+automatically. Keep each document durable and compact: store facts, source URLs,
+and decisions, not full timelines or private data.
+
+Xquik is an independent third-party service. Not affiliated with X Corp.
+"Twitter" and "X" are trademarks of X Corp.

package.json

@@ -17,7 +17,7 @@
   "overrides": {
     "tar": "7.5.16"
   },
-  "description": "Membase plugin for OpenClaw — persistent long-term memory for AI agents",
+  "description": "Membase plugin for OpenClaw - persistent long-term memory for AI agents",
   "files": [
     "dist/",
     "src/client.ts",
@@ -33,6 +33,7 @@
     "src/utils.ts",
     "src/wiki-project.ts",
     "skills/",
+    "examples/",
     "index.ts",
     "openclaw.plugin.json"
   ],

src/hooks/capture.ts

@@ -29,6 +29,16 @@ const pendingDocumentBuffers = new Map<
 const silenceTimers = new Map<string, ReturnType<typeof setTimeout>>();
 let flushSequence = 0;
 
+export function clearCaptureState(): void {
+  for (const timer of silenceTimers.values()) {
+    clearTimeout(timer);
+  }
+  messageBuffers.clear();
+  pendingDocumentBuffers.clear();
+  silenceTimers.clear();
+  flushSequence = 0;
+}
+
 function getChannelKey(event: Record<string, unknown>): string {
   // Prefer the top-level sessionKey from newer OpenClaw event payloads.
   // Fall back to the legacy session object shape, then to "default".

src/membase-tools.test.ts

@@ -7,7 +7,11 @@ import {
   resolveTokenFilePath,
 } from "./config";
 import { formatBundles, formatWikiDocument } from "./format";
-import { flushAllBuffers, registerCaptureHook } from "./hooks/capture";
+import {
+  clearCaptureState,
+  flushAllBuffers,
+  registerCaptureHook,
+} from "./hooks/capture";
 import { registerAddWikiTool } from "./tools/add-wiki";
 import {
   buildCurrentDateText,
@@ -23,6 +27,7 @@ const originalFetch = globalThis.fetch;
 
 afterEach(() => {
   globalThis.fetch = originalFetch;
+  clearCaptureState();
 });
 
 function makeApi(): { api: OpenClawPluginApi; tools: ToolDefinition[] } {

src/tools/add-wiki.ts

@@ -62,9 +62,14 @@ export function registerAddWikiTool(
           return await toolResponse(`Add wiki failed: ${projectInput.error}`);
         }
         if (looksSensitive(params.content)) {
-          return await toolResponse(
-            "Add wiki failed: content appears to contain secrets or private credentials. Redact it before saving.",
-          );
+          return {
+            content: [
+              {
+                type: "text" as const,
+                text: "Add wiki failed: content appears to contain secrets or private credentials. Redact it before saving.",
+              },
+            ],
+          };
         }
         const doc = await client.createWikiDocument(
           params.title,

src/tools/update-wiki.ts

@@ -78,9 +78,14 @@ export function registerUpdateWikiTool(
           typeof params.content === "string" &&
           looksSensitive(params.content)
         ) {
-          return await toolResponse(
-            "Update wiki failed: content appears to contain secrets or private credentials. Redact it before saving.",
-          );
+          return {
+            content: [
+              {
+                type: "text" as const,
+                text: "Update wiki failed: content appears to contain secrets or private credentials. Redact it before saving.",
+              },
+            ],
+          };
         }
 
         const updates: Parameters<MembaseClient["updateWikiDocument"]>[1] = {