Capawesome Cloud

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only Capawesome Cloud skill for Capacitor builds, live updates, and app-store publishing; its sensitive workflows are disclosed and purpose-aligned, but users must handle credentials and deployment commands carefully.

Install this skill only if you intend to manage a Capacitor app with Capawesome Cloud. Treat Apple, Google, CI, keystore, certificate, service-account, and live-update signing materials as production secrets: use secret managers or masked CI variables, avoid pasting secrets into chat or shell history, and rotate anything exposed. Before deployment, deletion, live-update upload, or any command using --yes, explicitly confirm the app, channel, destination, track, build, and whether the action affects production users.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (15)

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill instructs users to authenticate with `npx @capawesome/cli login --token <token>` and later references credentials for Apple and Google publishing, but it does not warn that these tokens and credentials are sensitive secrets that must not be pasted into chat, committed to source control, or logged in CI output. In an agent setting, omission of secret-handling guidance increases the chance that users disclose long-lived credentials to the model or store them insecurely, which could enable unauthorized builds, deployments, or store publishing.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: This section instructs the agent/user to handle highly sensitive publishing credentials such as Apple API keys, app-specific passwords, and Google service account JSON files, but provides no warning to avoid exposing them in chat, logs, shell history, or source control. Because these credentials can enable store submissions and other account actions, mishandling them can lead to credential leakage or unauthorized releases.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The build-and-deploy flow states that a build can be created and automatically deployed to a configured destination, but it does not warn that this may submit to TestFlight, the App Store, or Google Play with production consequences. In an agent skill, omission of an explicit confirmation step increases the risk of accidental publishing or release to the wrong destination/track.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The documentation shows keystore and key passwords passed directly as CLI arguments, which can expose secrets through shell history, terminal logs, CI job output, and process listings on multi-user systems. In a build/signing context these credentials protect Android release keys, so leakage can enable unauthorized signing or compromise of the app release pipeline.

Missing User Warnings

Medium

Confidence: 78% confidence
Finding: The delete example uses a force-like confirmation bypass (`--yes`) without warning that deleting a signing certificate may be irreversible in the service and can disrupt builds or release operations. In the context of Android signing material, accidental deletion can cause operational outages and recovery difficulties, especially if the original keystore is not safely retained.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The documentation instructs users to pass the certificate password directly on the command line via `--password <CERTIFICATE_PASSWORD>`. Secrets provided as CLI arguments can be exposed through shell history, process listings, CI job logs, and audit tooling, which is especially sensitive here because the password protects an iOS signing certificate used to produce trusted app binaries.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The documentation shows token-based authentication via a command-line flag, which commonly exposes secrets through shell history, process listings, CI job logs, and copied command snippets. In a CI/CD-oriented skill, this increases the likelihood that users will handle API tokens unsafely even if the CLI technically supports the option.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The certificate commands document uploading keystores, provisioning profiles, and supplying passwords directly on the command line without any security guidance. This can lead to leakage of signing materials and passwords via shell history, terminal recordings, CI logs, or insecure storage, which would allow unauthorized signing of malicious app builds.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The environment-setting commands allow secrets via key=value arguments and .env files but provide no warning that these inputs may contain credentials or API keys. Users may commit .env files, expose secrets in shell history, or log them in automation, causing credential disclosure and downstream compromise.

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The destination configuration examples accept Google service account keys and Apple publishing credentials without any warning that these are privileged deployment secrets. Exposure of these credentials could let an attacker publish malicious releases, alter store configuration, or access developer accounts.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The live update commands reference private signing keys and immediate deployment actions without warning about key custody or the risk of shipping unreviewed bundles. In an OTA update workflow, compromise of the signing key or careless deployment can enable malicious code delivery to production devices outside normal app store review cycles.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The document instructs users to generate and download a Google service account JSON key, which is a highly sensitive long-lived credential that can enable Play Store release actions if exposed. Although the file does say to save it securely, it does not clearly warn that the key must never be committed to source control, shared in chat, or stored insecurely, which increases the risk of credential leakage during CI/CD setup.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The documentation instructs users to pass a private signing key on the command line without warning that CLI arguments may be exposed through shell history, process listings, CI job logs, or copied scripts. If the private key is disclosed, an attacker could sign malicious live-update bundles that appear trusted by the app, undermining update integrity.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: Registering a self-hosted bundle URL without trust and integrity guidance can lead users to fetch update artifacts from infrastructure that is improperly secured, mutable, or controlled by an attacker. In a live-update system, compromise of the hosting origin or transport path could enable delivery of malicious application code to devices.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The Bitbucket Pipelines example tells users to store CAPAWESOME_CLOUD_TOKEN as a repository variable, but does not specify that it must be secured or masked. In Bitbucket, plain repository variables may be more broadly exposed than secured variables, increasing the risk that the token could be revealed in logs, pipeline output, or to users with repository settings access, which could allow unauthorized live update uploads.

VirusTotal

67/67 vendors flagged this skill as clean.

View on VirusTotal