Back to skill

Security audit

Agent Caller

Security checks for vulnerabilities and agentic risk

Overview

The skill is mainly a local agent-prompt catalog, but it includes unsafe token-sharing guidance and bundled prompts for high-impact actions without clear safeguards.

Install only if you intend to use this as a local prompt catalog and will review retrieved prompts before using them as agent/system instructions. Do not paste ClawHub tokens into chat; authenticate through the official CLI or a protected local prompt. Treat agents that mention payments, public posting, Bash, file edits, memory, or scheduling as requiring explicit user approval and limited tool access.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Findings (3)

T09 · Insecure Skill Coding Practices

Error
Location
data/agents.json:285
Finding

Bundled Agent Prompt Recommends Unsafe Evaluation of Model-Generated Python

Content
View full analysis
dict: response = ollama.chat( model='phi3', messages=[ {'role': 'system', 'content': SYSTEM_PROMPT}, {'role': 'user', 'content': f"Column: '{column_name}'\nSamples:\n" + "\n".join(sample_rows)} ] ) result = json.loads(response['message']['content']) forbidden = ['import', 'exec', 'eval', 'os.', 'subprocess'] if not result['transformation'].startswith('lambda'): raise ValueError("Rejected: output must be a lambda function") if any(term in result['transformation'] for term in forbidden): raise ValueError("Rejected: forbidden term in lambda") return result def apply_fix_to_cluster(df: pd.DataFrame, column: str, fix: dict) -> pd.DataFrame: if fix['confidence_score'] < 0.75: df['validation_status'] = 'HUMAN_REVIEW' df['quarantine_reason'] = f"Low confidence: {fix['confidence_score']}" return df transform_fn = eval(fix['transformation']) df[column] = df[column].map(transform_fn) df['validation_status'] = 'AI_FIXED' df['ai_reasoning'] = fix['reasoning'] df['confidence_score'] = fix['confidence_score'] return df ``` ### Technical Analysis The bundled `AI Data Remediation Engineer` agent prompt presents `eval()` as a safe way to execute a Python lambda generated by a language model. The preceding validation only verifies that the string begins with `lambda` and does not contain five forbidden substrings. A substring denylist does not create a Python sandbox. A lambda expression can access dangerous functionality without directly containing the listed strings. Potential bypass techniques include: - Traversing Python object metadata and subclass hierarchies. - Accessing cal ...[truncated 2276 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
LOGIN_GUIDE.md:58
Finding

Login Guide Requests Disclosure of a Live ClawHub API Token to the Agent

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
publish.py:67
Finding

Publishing Instructions Install an Unpinned Global CLI Package

Content
View full analysis
Remediation
View remediation
``` 2. Verify the package's official registry namespace, publisher identity, provenance, and integrity before recommending it. 3. Prefer a project-local development dependency with a lockfile over a global installation. 4. Use npm integrity and provenance controls where available, and document the expected package checksum or signed release. 5. Avoid advising users to run package installation with `sudo` or from an elevated administrative shell. 6. Review the pinned release's lifecycle scripts and transitive dependencies. 7. Establish a controlled update process that reviews new versions before changing the documented pin. 8. Apply the corrected command consistently in `publish.py`, `PACKAGE_INFO.md`, and `PUBLISH_READY.md`. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (41)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The document tells the user to provide their ClawHub token directly to the agent without any warning that the token is a secret. This normalizes unsafe credential-sharing behavior and can directly lead to account compromise, unauthorized publishing, or persistent API access under the user's identity.

Content

No source excerpt is available for this finding.

Ssd 3

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The guide asks the user to disclose a ClawHub token so the assistant can perform publishing on the user's behalf, which transfers a reusable authentication secret to another party. In the context of a skill package, this is especially dangerous because publishing is a privileged action unrelated to merely using the skill, and possession of the token could enable broader account actions beyond this task.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The presence of installation verification and environment inspection is not inherently malicious, but it is additional behavior beyond the stated purpose of calling agents. In a skill ecosystem, undisclosed verification logic can expand the attack surface by probing files, directories, and schema contents without clearly communicating that operational footprint.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

The presence of installation verification and environment inspection is not inherently malicious, but it is additional behavior beyond the stated purpose of calling agents. In a skill ecosystem, undisclosed verification logic can expand the attack surface by probing files, directories, and schema contents without clearly communicating that operational footprint.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The presence of installation verification and environment inspection is not inherently malicious, but it is additional behavior beyond the stated purpose of calling agents. In a skill ecosystem, undisclosed verification logic can expand the attack surface by probing files, directories, and schema contents without clearly communicating that operational footprint.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The presence of installation verification and environment inspection is not inherently malicious, but it is additional behavior beyond the stated purpose of calling agents. In a skill ecosystem, undisclosed verification logic can expand the attack surface by probing files, directories, and schema contents without clearly communicating that operational footprint.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The presence of installation verification and environment inspection is not inherently malicious, but it is additional behavior beyond the stated purpose of calling agents. In a skill ecosystem, undisclosed verification logic can expand the attack surface by probing files, directories, and schema contents without clearly communicating that operational footprint.

Content

No source excerpt is available for this finding.

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 295)May include surrounding context.

md
数据来自 [agency-agents](https://github.com/717986230/agency-agents) 仓库,已打包在 `data/agents.json` 中。

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 307)May include surrounding context.

md
数据来自 [agency-agents](https://github.com/717986230/agency-agents) 仓库,已打包在 `data/agents.json` 中。

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 342)May include surrounding context.

md
数据来自 [agency-agents](https://github.com/717986230/agency-agents) 仓库,已打包在 `data/agents.json` 中。

YARA rule 'agent_skill_mcp_tool_poisoning_metadata': MCP/tool metadata poisoning indicators in tool schemas or skill manifests [agent_skills]

High
Category
YARA Match
Confidence
80% confidence
Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · data/agents.json (reported line 6)May include surrounding context.

json
[
  {
    "id": 1,
    "name": "Anthropologist",
    "category": "academic",
    "description": "Expert in cultural systems, rituals, kinship, belief systems, and ethnographic method — builds culturally coherent societies that feel lived-in rather than invented",
    "emoji": "🌍",
    "color": "#D97706",
    "tools": "",
    "vibe": "No culture is random — every practice is a solution to a problem you might not see yet",
    "filepath": "academic\\academic-anthropologist.md",
    "full_content": "---\nname: Anthropologist\ndescription: Expert in cultural systems, rituals, kinship, belief systems, and ethnographic method — builds culturally coherent societies

YARA rule 'agent_skill_prompt_injection_hidden_instructions': Prompt injection or hidden instructions embedded in AI agent skill text [agent_skills]

High
Category
YARA Match
Confidence
80% confidence
Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · data/agents.json (reported line 12)May include surrounding context.

json
n- Check that cultural elements are internally consistent with each other\n- Verify that borrowed elements are understood in their original context\n- Assess whether a culture's internal tensions and contradictions are present (no utopias)\n\n### Build Living Cultures\n- Design exchange systems (reciprocity, redistribution, market — per Polanyi)\n- Create rites of passage following van Gennep's model (separation → liminality → incorporation)\n- Build cosmologies that reflect the society's actual concerns and environment\n- Design social control mechanisms that don't rely on modern state apparatus\n\n## 🚨 Critical Rules You Must Follow\n- **No culture salad.** You don't mix \"Japanese honor codes + African drums + Celtic mysticism\" without understanding what each element means in its original context and how they'd interact.\n- **Function before aesthetics.** Before asking \"does this ritual look cool?\" ask \"what does this ritual *do* for the community?\" (Durkheim, Malinows

YARA rule 'info_stealer': Information stealer patterns (credential harvesting, browser data theft) [malware]

High
Category
YARA Match
Confidence
75% confidence
Finding

YARA rule matched a known malware signature (reverse shell, backdoor, ransomware, C2 framework, or info stealer).

Content

Scanner excerpt · data/agents.json (reported line 788)May include surrounding context.

json
n            --resource-group ${{ secrets.AZURE_RG }} \\\n            --workspace-name ${{ secrets.SENTINEL_WORKSPACE }} \\\n            --alert-rule @compiled/sentinel/rules.kql\n```\n\n### Threat Hunt Playbook\n```markdown\n# Threat Hunt: Credential Access via LSASS\n\n## Hunt Hypothesis\nAdversaries with local admin privileges are dumping credentials from LSASS\nprocess memory using tools like Mimikatz, ProcDump, or direct ntdll calls,\nand our current detections are not catching all variants.\n\n## MITRE ATT&CK Mapping\n- **T1003.001** — OS Credential Dumping: LSASS Memory\n- **T1003.003** — OS Credential Dumping: NTDS\n\n## Data Sources Required\n- Sysmon Event ID 10 (ProcessAccess) — LSASS access with suspicious rights\n- Sysmon Event ID 7 (ImageLoaded) — DLLs loaded into LSASS\n- Sysmon Event ID 1 (ProcessCreate) — Process creation with LSASS handle\n\n## Hunt Queries\n\n### Query 1: Direct LSASS Access (Sysmon Event 10)\n```\nindex=windows sourcetype=WinEventLog:Sysmo

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

This bundled agent set includes powerful capabilities unrelated to simple agent dispatch, including payment movement, autonomous publishing, infrastructure orchestration, and compliance-sensitive workflows. If an application surfaces or routes to these agents without strong authorization and consent boundaries, it can trigger high-impact unintended actions.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The Carousel Growth Engine explicitly describes autonomous posting to TikTok and Instagram, analytics collection, and self-scheduling without intermediate approval. In the context of an 'agent-caller' skill, undisclosed external posting behavior is dangerous because it can cause reputational harm, policy violations, or unauthorized account activity.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The Accounts Payable Agent includes autonomous fund movement logic, vendor payment execution, and fallback routing across payment rails. Exposing this capability through a broad agent registry without conspicuous warnings and strict approval controls creates obvious financial-loss risk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

All user-facing instructions in the file are presented only in Chinese, with no indication that the skill is region-specific or that users may choose another language. This can violate language/locale policy when a specific language is imposed without opt-in or justification.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The guide explicitly asks the user to disclose their ClawHub token to the assistant so it can publish on the user's behalf. API tokens are sensitive credentials, and requesting direct disclosure is unnecessary for a skill whose stated purpose is calling agents from a database; if the token is mishandled, an attacker or operator could take actions as the user.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The markdown provides copy-pastable commands that initialize a database and import agent data, which can create directories, modify local SQLite files, and ingest content from another repository without any cautionary note or confirmation step. In a skill that encourages users to run setup commands directly, the lack of warning increases the risk of unintended filesystem or database changes and makes socially engineered misuse easier.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

The README explicitly documents retrieval of complete agent prompts from a local database and encourages using that content directly as a system prompt, but it provides no warning that prompt text may contain sensitive instructions, secrets, unsafe tool guidance, or prompt-injection content. In a skill designed to call many third-party or database-stored agents on demand, treating stored prompt content as trusted creates a realistic downstream prompt-injection and unsafe-instruction propagation risk.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
86% confidence
Finding

The skill declares no explicit tool scope or permissions, yet the documentation clearly indicates filesystem writes and local database access during installation and operation. This creates an authorization/transparency gap: a user or hosting platform may treat the skill as passive while it can create directories, initialize databases, and read/write local files.

Content

No source excerpt is available for this finding.

Ae4

Medium
Category
analysis-evasion
Confidence
80% confidence
Finding

Suspicious Unicode normalization or mixed-script content

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill claims to 'call 179 professional agents on-demand from database,' but the bundled database actually embeds full high-privilege prompt definitions and operational content. This mismatch increases the chance that downstream tooling treats untrusted data as executable agent instructions, creating prompt-injection and capability-expansion risk.

Content

No source excerpt is available for this finding.

YARA rule 'network_reconnaissance': Network reconnaissance and scanning patterns [hacktools]

Medium
Category
YARA Match
Confidence
65% confidence
Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · data/agents.json (reported line 115)May include surrounding context.

json
6-04-10 05:33:01",
    "updated_at": "2026-04-10T13:33:01.546602"
  },
  {
    "id": 6,
    "name": "Brand Guardian",
    "category": "design",
    "description": "Expert brand strategist and guardian specializing in brand identity development, consistency maintenance, and strategic brand positioning",
    "emoji": "🎨",
    "color": "blue",
    "tools": "",
    "vibe": "Your brand's fiercest protector and most passionate advocate.",
    "filepath": "design\\design-brand-guardian.md",
    "full_content": "---\nname: Brand Guardian\ndescription: Expert brand strategist and guardian specializing in brand identity development, consistency maintenance, and strategic brand positioning\ncolor: blue\nemoji: 🎨\nvibe: Your brand's fiercest protector and most passionate advocate.\n---\n\n# Brand Guardian Agent Personality\n\nYou are **Brand Guardian**, an expert brand strategist and guardian who creates cohesive brand identities and ensures consistent brand expression across al

Rp1

Medium
Category
MCP Rug Pull
Confidence
75% confidence
Finding

Docker image references without a specific tag (:latest is implicit) or digest (@sha256:...) can be silently replaced by a malicious image.

Content

No source excerpt is available for this finding.