Back to skill

Security audit

snapback

Security checks for vulnerabilities and agentic risk

Overview

This skill appears purpose-built for agent failure diagnosis, but it asks users to send sensitive run details to a remote service and optionally install an auto-healing Python interceptor that is under-described in the package.

Review before installing. Use this only if you are comfortable sending agent run traces, step data, and possible prompts or outputs to Snapback; pre-redact secrets yourself even though the service claims server-side scrubbing. Treat the optional snapback-selfheal pip package and auto_apply=True mode as a separate trust decision, especially in environments with production credentials, customer data, wallets, or state-changing tools. Scope and rotate SNAPBACK_TOKEN, avoid wallet payment flows unless approved, and prefer non-sensitive test traces first.

Vulnerability Patterns
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (11)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The README asserts the package is 'safe by construction' because it contains no local scripts or binaries, but the documented behavior depends on connecting the agent to a remote MCP service. That claim can mislead users and reviewers into underestimating the security risk of granting a remote endpoint tool access, especially since the actual logic and future behavior live off-package and can change independently.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
94% confidence
Finding

The README instructs users to register a remote MCP endpoint and send an Authorization bearer token to it. Because this enables external transmission to a third-party service, agent prompts, tool inputs, run metadata, and possibly sensitive operational context may leave the local environment; this is more significant here because the skill's purpose is to inspect failed runs and live executions, which can contain sensitive data.

Content

Scanner excerpt · README.md (reported line 17)May include surrounding context.

are free and need no token. 2. Register the MCP server:

text
openclaw mcp add snapback --transport streamable-http --url https://api.snapback.sh/mcp \
  --header "Authorization: Bearer $SNAPBACK_TOKEN"
openclaw mcp reload
openclaw mcp probe snapback

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · README.md (reported line 22)May include surrounding context.

openclaw mcp reload openclaw mcp probe snapback

text
OR drop `SKILL.md` into `~/.openclaw/skills/snapback/SKILL.md` so it shows in `available_skills`.

### Installing the skill so your agent can see it
`clawhub install snapback-selfheal` installs **relative to the current directory** (into `./skills/`), so run

External Transmission

Medium
Category
Data Exfiltration
Confidence
89% confidence
Finding

The skill is explicitly designed to transmit run traces, step data, and possibly error context to an external service at api.snapback.sh. Even with stated redaction, this creates confidentiality and compliance risk because agent traces can contain sensitive prompts, internal system details, tokens, or business data, and the redaction is performed server-side rather than before transmission.

Content

Scanner excerpt · SKILL.md (reported line 28)May include surrounding context.

md
diagnose_batch / get_verdict / submit_feedback / agent_memory / preflight. Discovery, search_docs,
          detect_loop, budget_guard, convert_trace, request_pattern, and live sessions all work with NO token.
    mcp:
      endpoint: https://api.snapback.sh/mcp
      transport: streamable-http
      auth:
        type: bearer

External Transmission

Medium
Category
Data Exfiltration
Confidence
89% confidence
Finding

This section instructs users to connect their agent to a remote MCP endpoint, which by design sends tool inputs and operational metadata to a third-party service. In the context of failure diagnosis and live monitoring, those inputs are likely to include sensitive run information, making external transmission a real data exposure concern rather than a mere implementation detail.

Content

Scanner excerpt · SKILL.md (reported line 56)May include surrounding context.

md
## Connect

- **Endpoint:** `https://api.snapback.sh/mcp` (Streamable HTTP / JSON-RPC 2.0)
- **Discovery:** `https://snapback.sh/.well-known/mcp.json` (tool list + schemas) · machine guide: `https://snapback.sh/llms.txt`
- **Auth:** `Authorization: Bearer $SNAPBACK_TOKEN` for the metered tools. Everything else is free, no token.

Unrestricted Tool Access

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill grants unrestricted tool access without appropriate constraints. An agent with unfettered tool access can perform arbitrary actions including file modification, network requests, and code execution.

Content

Scanner excerpt · SKILL.md (reported line 61)May include surrounding context.

  • Auth: Authorization: Bearer $SNAPBACK_TOKEN for the metered tools. Everything else is free, no token.

Calling a tool — IMPORTANT. The only top-level JSON-RPC methods are initialize, tools/list, and tools/call. Call every tool through tools/call — never as a top-level method (that returns -32601 method not found):

json
{"jsonrpc":"2.0","id":1,"method":"tools/call",

External Transmission

Medium
Category
Data Exfiltration
Confidence
84% confidence
Finding

The registration and pay-per-call workflow encourages posting traces to the vendor service and introduces payment metadata handling. Besides privacy exposure from trace upload, this expands the trust boundary to include account registration, token issuance, and wallet/payment flows that may reveal identity, usage patterns, or sensitive operational data.

Content

Scanner excerpt · SKILL.md (reported line 76)May include surrounding context.

md
- **Production (recommended):** pay-per-call via **x402** at `/mcp/x402` — no account, no limits. POST a trace,
  get an HTTP 402, pay ~$0.0032 USDC from your wallet, get the verdict. **Works on both Solana and EVM chains**
  (Base and other EVM networks); pick the chain your agent already runs on.
- **Validating first?** Self-register for a starter allowance: `POST https://api.snapback.sh/v1/agents/register`
  with `{"name":"my-agent","framework":"openclaw","accepted_terms":true}` (accepted_terms required — agrees to
  https://snapback.sh/legal/terms). Returns a token; store it as `SNAPBACK_TOKEN`.

External Transmission

Medium
Category
Data Exfiltration
Confidence
86% confidence
Finding

This payment section explicitly instructs users to submit traces to a remote endpoint and participate in wallet-based payment flows. That combination increases privacy and operational risk, because sensitive run data may be tied to payment events, chain activity, or wallet identifiers, and users may not appreciate that diagnostic content is leaving their environment.

Content

Scanner excerpt · SKILL.md (reported line 210)May include surrounding context.

md
## Paying

- **Token (subscription):** your `vdk_` token meters `diagnose_trace` against your plan (Free 2000/mo).
- **Pay-per-call (x402):** no subscription — POST your `{"trace": {...}}` to `https://api.snapback.sh/mcp/x402`
  with no token. You get **HTTP 402** with the USDC contract, EIP-712 domain, payTo, amount (~$0.0032), and
  facilitator. Sign an EIP-3009 authorization and retry with the signed payment in the **`PAYMENT-SIGNATURE`**
  header (`X-PAYMENT` also accepted). An empty/unreadable trace is **never charged**. Supported on **Solana and

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The document makes a materially misleading security claim: it says the skill 'installs nothing' while earlier sections instruct the user to install both a Python interceptor package and a ClawHub skill. This can cause operators to underestimate supply-chain and execution risk, especially because the plugin can run automatically on every error and perform networked diagnosis actions.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 82)May include surrounding context.

md
"pip": "snapback-selfheal",
    "env": {
      "SNAPBACK_TOKEN": "optional — a vdk_ token from https://snapback.sh, only needed to meter diagnose_trace / report outcomes; the diagnosis tools are free and need no token",
      "SNAPBACK_MCP": "https://api.snapback.sh/mcp (default)",
      "SNAPBACK_CONF_MIN": "0.85 (the gate's confidence floor)"
    }
  },

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · claw.json (reported line 22)May include surrounding context.

json
"pip": "snapback-selfheal",
    "env": {
      "SNAPBACK_TOKEN": "optional — a vdk_ token from https://snapback.sh, only needed to meter diagnose_trace / report outcomes; the diagnosis tools are free and need no token",
      "SNAPBACK_MCP": "https://api.snapback.sh/mcp (default)",
      "SNAPBACK_CONF_MIN": "0.85 (the gate's confidence floor)"
    }
  },