What makes research-assistant notable?

985 downloads. Trust evidence available

How should research-assistant be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

research-assistant

Knowledge builder that extracts entities, relationships, and key facts from web pages, documents, and files. Builds a searchable knowledge base with entity r... Skill: research-assistant Owner: jlacroix82 Summary: Knowledge builder that extracts entities, relationships, and key facts from web pages, documents, and files. Builds a searchable knowledge base with entity r... Tags: latest:1.1.5, security-fix:1.1.0 Version history: v1.1.5 | 2026-07-22T22:41:20.347Z | user v1.1.5: Security audit remediation — removed 'fully local' / 'zero external dependencies' claims, fixed email

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

985

Updated

Oct 11, 2026

Version

1.1.5

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 985 downloads reported by the source. Last updated 10/11/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 11, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 11, 2026
Adoption signal: 985 downloadsadoption · observed Oct 11, 2026
Latest release: 1.1.5release · observed Jul 22, 2026
Handshake status: UNKNOWNsecurity

What changed

Newest first. Only changes that point at a source are listed.

Release 1.1.5release
v1.1.5: Security audit remediation — removed 'fully local' / 'zero external dependencies' claims, fixed email extraction documentation contradictions across SKILL.md and README.md, added explicit outbound HTTP and persistent storage warnings, excluded test/ directories from published bundle
Observed Jul 22, 2026 · source

Install and run

Setup complexity: low.

clawhub skill install s175p518b8g47fx6r9zyvs95ks876t4t:agent-researcher

Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.
Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-jlacroix82-agent-researcher/snapshot"

Run-check

$0.02 USD

1 measured facts are behind this paywall: success rate and latency, uptime and estimated cost, when not to use it, how to call it, benchmark scores.

Agents pay $0.02 in USDC. A card payment is $0.50, the smallest a card allows.

Documentation

CLAWHUB

94,061 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: research-assistant
description: Knowledge builder that extracts entities, relationships, and key facts from web pages, documents, and files. Builds a searchable knowledge base with entity resolution and auto-summarization. Integrates with memory-router.
---

# Research Assistant ⚡

**Turn everything you read into a searchable knowledge base.**

## The Problem

Agents read web pages, docs, and files — but the knowledge disappears after the session. No persistent knowledge base, no entity tracking, no relationship mapping.

Research Assistant fixes this with one tool.

## ⚠️ Important Warnings

> **Read this section before using.** Research Assistant makes outbound HTTP requests to user-supplied URLs and writes all extracted content (entities, summaries, relations) to persistent local files under `memory/research/`. These are not optional behaviors and they cannot be disabled. Do not point this tool at confidential documents, internal network endpoints, or sources containing PII.

### Outbound Network Access
`--extract <url>` makes an unauthenticated HTTP/HTTPS request to any user-supplied URL (10-second timeout). **No URL allowlist, rate limit, or proxy is applied.** The target server will see your real IP address. There is no SSRF protection — internal/private network URLs (10.x, 192.168.x, localhost, etc.) are NOT blocked. Treat this as remote code-execution-grade network exposure: only point it at URLs you would type into a browser.

### Persistent Local Storage
All extracted entities, summaries, text content, and relationships are written to `memory/research/` and persist indefinitely:
- `knowledge-base.json` — every entry ever extracted
- `index.json` — searchable entity index
- `relations.json` — relationship graph

These files grow unbounded with every extraction. There is no retention limit, automatic redaction, or built-in deletion command. PII or sensitive content ingested from URLs and files is stored forever. Monitor with `--status` and manually clean the `memory/research/` directory if needed.

### PII in Extracted Data
Entity extraction captures proper nouns, URLs, numbers, and dates. **Email address extraction is intentionally disabled** — the `extractEntities()` function does not match email patterns (see line 90 of `research-assistant.js`: `// NOTE: Email addresses are intentionally NOT extracted to avoid PII leakage`). However, email addresses may still appear inside the `summary` text field (first 200 chars of source) or inside HTML content fetched from URLs. Do not rely on the email-extraction disablement as a complete PII filter — review the stored knowledge base before exporting or sharing it.

## Quick Start

### Extract from a file

```bash
node skills/research-assistant/research-assistant.js --extract /path/to/file.md
```

Extracts entities (proper nouns, numbers, URLs) and relationships, and creates a summary. Email address extraction is disabled — see "PII in Extracted Data" above.

### Extract from a URL

```b

README.md

# Research Assistant 🔬

**Knowledge builder that extracts entities, relationships, and key facts from web pages, documents, and files. Builds a searchable knowledge base with entity resolution and auto-summarization.**

> **Read the "Security & Privacy" section before using.** This tool makes outbound HTTP requests to user-supplied URLs and writes all extracted content to persistent local files. These behaviors cannot be disabled.

## Why Research Assistant?

Information is everywhere — web pages, documents, code files, emails. Research Assistant makes it *usable*:

- **Entity extraction** — Automatically finds proper names, dates, URLs, and key concepts
- **Searchable knowledge base** — Queries across everything you've processed
- **Knowledge graph** — Visualize entity relationships and connections
- **Domain filtering** — Focus on specific topics or sources
- **Auto-summarization** — Key entities and sources at a glance
- **Pure Node.js** — No npm runtime dependencies

## ⚠️ What This Tool Does (Honest Description)

- **Outbound HTTP**: `--extract <url>` fetches remote URLs. No allowlist, no rate limit, no proxy, no SSRF protection. Treat this as remote network exposure — only point it at URLs you would type into a browser.
- **Persistent local storage**: Every extraction (URL or local file) is written to `memory/research/` and kept forever. There is no automatic retention or deletion.
- **Email extraction is disabled** at the code level (see `extractEntities()` line 90), but emails can still appear inside fetched HTML or in source text — they are stored in the `summary` field.
- **No shell execution, no system calls, no subprocess spawning.** The only network I/O is `--extract <url>` via Node's built-in `http`/`https` modules.

---

## Installation

```bash
# Already included in OpenClaw workspace at skills/research-assistant/
# No npm install needed — pure Node.js
```

---

## Quick Start

```bash
# Extract entities from text
node skills/research-assistant/research-assistant.js --extract "John Doe joined the project on 2026-06-15"

# Search existing knowledge base
node skills/research-assistant/research-assistant.js --search "John Doe"

# Add to knowledge base
node skills/research-assistant/research-assistant.js --add "document-id" --source "report.md"

# Show knowledge graph
node skills/research-assistant/research-assistant.js --graph

# Summarize knowledge base
node skills/research-assistant/research-assistant.js --summarize
```

---

## Commands Reference

### Entity Extraction

```bash
node skills/research-assistant/research-assistant.js --extract <text>
```

Extracts from unstructured text:
- **Proper nouns** — Names, places, organizations
- **Dates** — ISO dates, partial dates, date ranges
- **URLs** — http/https links
- **Numbers** — Counts, IDs, versions
- **Key phrases** — Multi-word concepts (via NLP-like heuristics)

Email address extraction is **disabled** in code. Emails may still appear in the stored `summary` field (first 200

_meta.json

{
  "ownerId": "kn7b6eyf5vc7khg5fr63pjm8xd82qvw5",
  "slug": "agent-researcher",
  "version": "1.1.5",
  "publishedAt": 1784760080347
}

skill-card.md

## Description:

Knowledge builder that extracts entities, relationships, and key facts from web pages, documents, and files into a searchable knowledge base with entity resolution and auto-summarization.

This skill is ready for commercial/non-commercial use.

## Publisher:

[jlacroix82](https://clawhub.ai/user/jlacroix82)

### License/Terms of Use:

MIT-0

## Use Case:

Developers and agent operators use this skill to extract entities, relationships, summaries, and key facts from web pages and local documents, then search or summarize a persistent local research knowledge base.

### Deployment Geography for Use:

Global

## Known Risks and Mitigations:

Risk: URL extraction can contact arbitrary user-supplied HTTP or HTTPS URLs, including internal or private-network destinations.

Mitigation: Use only trusted public URLs, avoid localhost and private-network targets, and add egress controls such as an allowlist or proxy before use in sensitive environments.

Risk: Extracted content, entities, summaries, and relationships are stored persistently and may include confidential information or PII.

Mitigation: Do not process confidential or PII-bearing sources unless retention, deletion, and redaction controls are added; manually review and clean memory/research/ as needed.

Risk: Local file and directory extraction can read broad workspace inputs and add them to the persistent knowledge base.

Mitigation: Run the skill from a scoped workspace and provide only approved files or directories for extraction.

## Reference(s):

- [ClawHub skill page](https://clawhub.ai/jlacroix82/skills/agent-researcher)
- [Publisher profile](https://clawhub.ai/user/jlacroix82)

## Skill Output:

**Output Type(s):** [text, shell commands, JSON files, guidance]

**Output Format:** [Markdown guidance with inline shell commands, CLI text output, and persistent local JSON knowledge-base files.]

**Output Parameters:** [1D]

**Other Properties Related to Output:** [Writes persistent local state under memory/research/; URL extraction can make outbound HTTP requests to user-supplied URLs.]

## Skill Version(s):

1.1.5 (source: server release metadata and clawhub.yaml)

## Ethical Considerations:

Users should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.

clawhub.yaml

slug: agent-researcher
name: research-assistant
owner: jlacroix82
description: Web search, content aggregation, and research assistance for agent-native skills. ⚠️ Persists extracted content and relationships to disk. Fetches arbitrary URLs. Review privacy and network safety before installing.
version: 1.1.5
type: skill
author: OpenClaw
license: MIT
keywords:
  - research
  - search
  - web
  - aggregation
  - agent
entry: SKILL.md
dependencies: []

Github ReposUpdated 3d agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.

MCPOPENCLAW

Github ReposUpdated 7mo agoRank 70

CopilotKit

The Frontend for Agents & Generative UI. React + Angular