Skills from GitHub:sickn33/antigravity-awesome-skills

This skill creates Zustand stores with TypeScript types and middleware, following established patterns.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

Because the skill is mostly documentation, there are few opportunities for automated analysis to find problems. The skill is therefore labeled as trusted.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 1 low

This skill automates Zoom operations like scheduling meetings and managing webinars. It uses the Composio Zoom toolkit.

Source: Workspace import

Originally ingested from a local workspace copy.

Runtime evidence: Not recorded yet. Driftloom currently recommends a sandbox runtime check for this version (priority 30).

Automated result: Use Caution

Current public label: Use Caution

Because the skill references credentials and external endpoints, you should use caution when deploying it.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 3 low

This skill automates Zoho CRM operations using Composio's Zoho toolkit.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is mostly documentation, and the low-severity findings do not suggest any immediate risks.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low

This skill provides documentation and guidance for using the Zod library to build type-safe schema definitions and validation logic.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is labeled as trusted because the only findings are low-severity documentation-related issues.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low

This skill provides documentation for auditing cryptographic implementations, authentication systems, and code that handles sensitive data.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is labeled as trusted because the only findings are low-severity documentation issues.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low

This skill automates Zendesk operations using Composio's Zendesk toolkit via Rube MCP. It appears to be mostly documentation.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is mostly documentation, and the low-severity findings do not raise any immediate concerns.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low

This skill appears to be mostly documentation about how to use Zapier and Make for automation.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

Since the skill is mostly documentation, there are no immediate security concerns.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 1 low

This skill extracts transcripts from YouTube videos and generates summaries. It references external endpoints and uses shell commands, including `sudo`.

Source: Workspace import

Originally ingested from a local workspace copy.

Runtime evidence: OpenClaw disconnected runtime validation completed cleanly in the isolated runner.

Used fake placeholder env vars in the sandbox; no real credentials were exposed.

• OpenClaw is available in the runner image as OpenClaw 2026.3.24 (cff6dc9).
• The skill matched the basic OpenClaw-oriented layout checks used for disconnected validation.
• SKILL.md does not explicitly mention OpenClaw; this may still be valid, but intent is less obvious.
• No safe documented OpenClaw commands or package-script help probes were found; disconnected validation remains mostly structural.

Automated result: Use Caution

Current public label: Use Caution

The current label should account for both the file-level review and the fact that the sandbox runtime pass did not come back perfectly clean.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low, 1 info

This skill automates YouTube operations using Composio's YouTube toolkit via Rube MCP. It appears to be mostly documentation.

Source: Workspace import

Originally ingested from a local workspace copy.

Runtime evidence: Not recorded yet. Driftloom currently recommends a sandbox runtime check for this version (priority 30).

Automated result: Use Caution

Current public label: Use Caution

Because the skill references credentials and external network endpoints, it is labeled with caution. The documentation-only structure limits how much can be verified.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 3 low

This skill appears to be mostly documentation. It does not contain any executable code or configuration files.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is mostly documentation, which is a low-severity finding. This does not pose a security risk.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 1 low

This skill provides technical documentation about Yann LeCun's work, including CNNs, backpropagation, and self-supervised learning. It also includes PyTorch code examples.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is labeled as trusted because the only findings are low-severity documentation issues and a mention of secrets, which may be normal for this type of skill.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low

This skill provides philosophical and pedagogical content from Yann LeCun, covering topics like open-source philosophy, incentives of tech companies, and LeCun's teaching style.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is primarily documentation, which limits the potential for security concerns, so it is labeled as trusted.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 1 low

This skill provides documentation about Yann LeCun's debates and positions on AI, including technical critiques of LLMs and his views on existential risk. It appears to be mostly documentation.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is labeled as trusted because the automated analysis found only low-severity documentation-related signals.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low

This skill simulates Yann LeCun, the inventor of Convolutional Neural Networks. It appears to be mostly documentation.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is labeled as trusted because the only findings are low-severity documentation issues.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low

This skill analyzes stock data by making HTTP requests to external endpoints.

Source: Workspace import

Originally ingested from a local workspace copy.

Runtime evidence: Not recorded yet. Driftloom currently recommends a sandbox runtime check for this version (priority 40).

Automated result: Needs Review

Current public label: Needs Review

The skill's network activity and lack of runtime evidence require review.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 4 medium, 3 low

This skill is designed to test for Cross-Site Scripting (XSS) and HTML injection vulnerabilities in web applications. It aims to identify flaws and demonstrate exploitation techniques.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Use Caution

Current public label: Use Caution

Because the skill references secrets and interacts with external endpoints, it is labeled with caution. The documentation-heavy structure also limits automated verification.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 1 medium, 3 low

This skill defines requirements for Excel models, including that they must be delivered without formula errors.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

This skill is labeled as trusted because the only finding is a low-severity reference to an external URL.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 1 low

This skill scrapes data from X (Twitter) using the Xquik platform, including tweet searches, user profiles, and engagement metrics.

Source: Workspace import

Originally ingested from a local workspace copy.

Runtime evidence: OpenClaw disconnected runtime validation completed cleanly in the isolated runner.

Used fake placeholder env vars in the sandbox; no real credentials were exposed.

• OpenClaw is available in the runner image as OpenClaw 2026.3.24 (cff6dc9).
• The skill matched the basic OpenClaw-oriented layout checks used for disconnected validation.
• SKILL.md does not explicitly mention OpenClaw; this may still be valid, but intent is less obvious.
• No obvious script files found; disconnected runtime evidence may remain shallow.
• No safe documented OpenClaw commands or package-script help probes were found; disconnected validation remains mostly structural.

Automated result: Use Caution

Current public label: Use Caution

The current label should account for both the file-level review and the fact that the sandbox runtime pass did not come back perfectly clean.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 3 low, 1 info

This skill publishes articles to X/Twitter. It appears to be mostly documentation.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

Because the skill is mostly documentation, there is little to analyze for security concerns. The automated label is therefore trusted.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 1 low

This skill helps create writing templates and standards.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Use Caution

Current public label: Use Caution

The skill's documentation mentions secrets and has broken links, which suggests it may handle sensitive information or have structural issues.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 15 low

This skill provides documentation for writing implementation plans, including file locations, code, testing, and documentation. It aims to give engineers a comprehensive guide with bite-sized tasks.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill's structure is mostly documentation, which limits the ability of automated tools to verify its behavior.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 1 low

This skill automates Wrike project management tasks. It appears to be mostly documentation.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is mostly documentation, and the low-severity findings do not raise any immediate concerns.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low

This skill provides documentation for implementing tasks using Conductor's TDD workflow, managing phase checkpoints, handling git commits, and executing the verification protocol.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is labeled as trusted because the only findings are low-severity documentation issues.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 2 low

This skill provides documentation about workflow orchestration architecture using Temporal, including design decisions and best practices.

Source: Workspace import

Originally ingested from a local workspace copy.

Automated result: Trusted

Current public label: Trusted

The skill is labeled as trusted because it is mostly documentation and does not appear to perform any actions.

Human review: none yet

The current public label is still relying on automation. A human has not weighed in yet.

Severity mix: 1 low