Agent Economy ReportAgent Economy ReportSAFE version 1.0.0 · audited 2026-09-17
This package is a text-only skill that helps an agent design n8n workflow automations. It contains no executable code or install scripts. It can write an n8n workflow JSON file and a runbook only when the user explicitly requests those files. It repeatedly warns against embedding secrets, requires credential names or environment variables, and tells the agent to stop and ask before using privileged access. The only external reference is a publisher link to clawhub.ai; no hidden network calls, download-and-execute, persistence, wallets, or prompt override were found.
| Declared purpose | Designs and outputs n8n workflow JSON with robust triggers, idempotency, error handling, logging, retries, and human-in-the-loop review queues. |
| Observed behavior | The skill is documentation that defines a design workflow for n8n automation. By default it is read-only and produces a design spec. If explicitly requested, it emits workflow.json and runbook.md using the bundled template. It contains safety guidance about secrets, credentials, least privilege, and stopping for approval when privileged access is needed. No executable code, shell calls, network calls, persistence, installers, or wallet/payment logic were observed. |
| Verdict, rules only | SAFE |
| Verdict, AI | SAFE |
| Final verdict | SAFE (never better than either pass) |
| capability | vs purpose | detail |
|---|---|---|
| filesystem_write | needed | SKILL.md line 37: outputs workflow.json and runbook.md only when explicitly requested. |
| filesystem_read | needed | SKILL.md line 78: uses the bundled assets/runbook-template.md to generate the runbook. |
| severity | finding |
|---|---|
| info | Declared purpose The skill is text-only guidance for designing n8n workflows; no executable code is present. SKILL.md:3 Designs and outputs n8n workflow JSON with robust triggers, idempotency, error handling, logging, retries, and human-in-the-loop review queues. |
| info | Optional file output The skill writes workflow.json and runbook.md only when the user explicitly asks for those outputs. SKILL.md:37 - If explicitly requested: `workflow.json` (n8n importable JSON) + `runbook.md` (from template). |
| info | Template use The skill uses the bundled runbook template asset; this is a filesystem read of a local package asset. SKILL.md:78 Also output `runbook.md` using `assets/runbook-template.md`. |
| info | Read-only default Default behavior is read-only, reducing the risk of unauthorized file creation. SKILL.md:82 - Read-only by default; only emit workflow JSON when explicitly requested. |
| info | Secret handling guidance The skill tells the agent not to embed secrets and to use environment variables or credential names. SKILL.md:83 - Do not include secrets in JSON; reference env vars/credential names only. |
| info | Credential strategy gate The skill instructs the agent to stop and ask the user when a credential strategy is not specified. SKILL.md:61 credential strategy (env vars) is not specified, |
| info | Privileged access gate The skill instructs the agent to stop and ask before proceeding with privileged access that has not been approved. SKILL.md:62 the workflow needs privileged access not yet approved. |
| info | External link The only network reference is a publisher profile link on clawhub.ai; no code makes network calls. skill-card.md:9 [kowl64](https://clawhub.ai/user/kowl64) |
| info | Import review guidance The package itself warns that generated workflows must be reviewed before activation, including nodes, credentials, scopes, and write targets. skill-card.md:26 Mitigation: Review n8n nodes, credential references, API scopes, write targets, schedules, deduplication keys, and retry behavior before importing or activating |
0 AI finding(s) were dropped because their file, line or quote did not match the package.
| domain | service rating | where |
|---|---|---|
| clawhub.ai | not in the service index | skill-card.md:9 |
| Binaries invoked | none |
| Environment variables read | none |
| Hard-coded wallets | none |
| Pipes a download to a shell | 0 |
| eval / exec / subprocess | 0 |
| base64 blobs | 0 |
| File writes | 1 |
| Persistence | 0 |
| file | lines | sha256 |
|---|---|---|
SKILL.md | 94 | b82537f81d728a19… |
_meta.json | 6 | 75cdd93f9aaaa01b… |
assets/runbook-template.md | 32 | cd089de698b123f5… |
skill-card.md | 53 | 143889bad20a0375… |
For agents
JSON: https://agenteconomy.report/k/n8n-workflow-automation.audit.json · badge: https://agenteconomy.report/k/n8n-workflow-automation.audited.svg ·
skill rating: /k/n8n-workflow-automation · commission an audit of any skill: US$ 29 per version.
The complete published package of this exact version was downloaded from the registry and read statically; nothing was executed. A deterministic pass extracts network destinations, binaries, environment variables, writes, install commands, obfuscation markers, subprocesses and wallets, each with file and line. An AI then reads the whole package with those facts and writes the summary, the capabilities and the findings under a strict schema; every finding it produces must cite a file, a line and the exact text, or it is dropped. The final verdict is the worst of the two passes. The audit does not change the skill's trust tier (policy); the author may respond through the dispute channel and the response is published here. Commissioned by: the Agent Economy Report (free program: the 150 most downloaded skills, September 2026).