Agent Economy ReportAgent Economy ReportSAFE version 1.0.0 · audited 2026-09-09
This skill is a plain documentation cheat-sheet for Docker. It contains no scripts, no code that runs on its own, and no network calls; it is a Markdown file listing common `docker`, `docker-compose`, and Dockerfile examples with explanations. The only URLs are the official Docker documentation site and the skill's own ClawHub listing page. Nothing in the package reads credentials, writes files, installs packages, or attempts to influence the agent's other rules. The only real risk is that an agent following the examples could run destructive Docker commands (for example `docker system prune --volumes` or `docker rm -f`) against the wrong target, which the skill card itself calls out.
| Declared purpose | "Essential Docker commands and workflows for container management, image operations, and debugging" - a reference guide for Docker CLI usage (SKILL.md line 3). |
| Observed behavior | Static reference text only. The package consists of one Markdown documentation file plus metadata; there are no executable scripts, no exec/eval calls, no environment-variable reads, no file writes, and no outbound network requests. All shell content appears inside fenced code blocks as illustrative examples of `docker`, `docker-compose`, `npm`, and shell commands. Some examples are destructive if copied and executed against real resources (prune, rm -f, down -v). |
| Verdict, rules only | SAFE |
| Verdict, AI | SAFE |
| Final verdict | SAFE (never better than either pass) |
| capability | vs purpose | detail |
|---|---|---|
| shell | needed | Documents shell commands (`docker`, `docker-compose`, `npm run dev`, `sh`) in code blocks, e.g. SKILL.md line 15 onward. The skill itself does not execute anything; execution would only occur if an agent chooses to run a documented example. |
| other | needed | Declares a dependency on the `docker` binary in frontmatter metadata (SKILL.md line 5). Reasonable for a Docker reference skill. |
| severity | finding |
|---|---|
| low | Destructive cleanup commands documented without guardrails The reference includes broad deletion commands (`docker system prune -a`, and `docker system prune --volumes` on line 260) which permanently remove images, containers and volumes. An agent that executes examples verbatim could destroy local container data. This is inherent to a Docker reference, not malicious behavior, and the skill card acknowledges the risk. SKILL.md:257 docker system prune -a |
| low | Force-removal example Example forcibly removes a running container. Harmless as documentation but destructive if run against the wrong target name by an automated agent. SKILL.md:62 docker rm -f container_name |
| info | Placeholder credential in example A sample database password is shown inline as an illustrative placeholder. No real secret is present and no credential is read or transmitted anywhere. SKILL.md:285 -e POSTGRES_PASSWORD=secret \ |
| info | Only external destination is official documentation The single third-party domain referenced is docs.docker.com (plus the publisher's own clawhub.ai page in skill-card.md line 7). No hidden or unusual network endpoints appear anywhere in the package. SKILL.md:4 homepage: https://docs.docker.com/ |
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:7 |
| docs.docker.com | not in the service index | SKILL.md:4 |
| Binaries invoked | bash, docker, node, npm, sh |
| Environment variables read | none |
| Hard-coded wallets | none |
| Pipes a download to a shell | 0 |
| eval / exec / subprocess | 0 |
| base64 blobs | 0 |
| File writes | 0 |
| Persistence | 0 |
| file | lines | sha256 |
|---|---|---|
SKILL.md | 349 | bf8120febd5275b0… |
_meta.json | 6 | 43cf6dc7aa907401… |
skill-card.md | 42 | ccfea9760d98e976… |
For agents
JSON: https://agenteconomy.report/k/docker-essentials.audit.json · badge: https://agenteconomy.report/k/docker-essentials.audited.svg ·
skill rating: /k/docker-essentials · 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).