DigitalOcean

v1.0.0

Manage DigitalOcean resources via API — Droplets (create/destroy/resize/power), DNS zones and records, Spaces (object storage), Databases, Firewalls, Load Balancers, Kubernetes, and account/billing info.

2· 1.1k·3 current·3 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
medium confidence
Purpose & Capability
The name, description, SKILL.md and the included Python script all align: the code calls the official DigitalOcean API and implements droplets, DNS, firewalls, spaces, databases, billing, SSH keys, images, etc. There are no obvious unrelated capabilities (no calls to non-DigitalOcean endpoints). However the registry metadata claims 'instruction-only' / no install but does include a code file, which is an inconsistency (minor but worth noting).
Instruction Scope
SKILL.md instructs the agent (and the user) to store a DigitalOcean API token at ~/.config/digitalocean/token and to call the bundled script or the DigitalOcean API directly. The runtime instructions and script stay within the stated domain (DigitalOcean API). The instructions do not ask the agent to read unrelated system files or exfiltrate data to unexpected endpoints.
Install Mechanism
There is no install spec (no packages downloaded), which is low risk. But the package is not truly 'instruction-only' — it bundles scripts/digitalocean.py that will be executed locally. That mismatch is informational: the code is included and will run when invoked, but nothing is installed automatically.
!
Credentials
SKILL.md and the script require a DigitalOcean API token stored at ~/.config/digitalocean/token, yet the registry metadata declares no primary credential, required env vars, or required config paths. This omission is an inconsistency that matters: the skill does require a secret (API token) with potentially broad privileges, but that requirement is not declared in the metadata. Users should treat this as suspicious until the credential requirements are explicit. Also SKILL.md suggests 'Spaces (requires spaces key)' which implies additional credentials; the metadata does not reflect that either.
Persistence & Privilege
The skill does not request always:true and does not modify other skills or system-wide settings. It only reads a token file from the user's home directory when invoked. Autonomous invocation (disable-model-invocation:false) is the platform default and not highlighted here; combined with the credential omission above it increases risk, but on its own this is normal.
What to consider before installing
This skill appears to implement a legitimate DigitalOcean CLI, but metadata and documentation disagree about credentials and packaging — the SKILL.md and script expect an API token at ~/.config/digitalocean/token while the registry metadata lists no required credentials or config paths. Before installing or using it: - Treat the repository as unverified (source/homepage unknown). Prefer official or well-known integrations when possible. - Inspect scripts/digitalocean.py yourself (you already have it) for any unexpected network addresses or file accesses (none were found beyond the token file and the DigitalOcean API in this scan). - Do not store a full-purpose account token unless you trust the skill; create a scoped API token with the minimal privileges needed (e.g., read-only or limited resource scopes), and rotate it after testing. - Consider placing the token in a secure location you control (or pass via a process environment variable) and verify the file permissions (SKILL.md recommends chmod 600). - If you plan to let an agent invoke this skill autonomously, be extra cautious: autonomous access plus an undeclared credential increases blast radius. Require explicit metadata that declares required credentials/config paths and confirm the token scope before enabling the skill. If the owner can update the registry metadata to declare the API token/config path (or provide a safer credential flow), that would resolve the main inconsistency and increase confidence.

Like a lobster shell, security has layers — review code before you run it.

latestvk974jwjpb198a2edvsgshjssb580x425

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments