agentic-awesome-skills@17.3.0
AAS Core: complete local skill discovery, agent-owned selection, stack validation, and planning, backed by 2,122+ agentic skills.
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThis package does not run on npm install, but the published CLI is built to copy a large mixed-risk skill catalog into AI-agent skill directories, and a binary PDF plus incomplete static coverage leave residual dual-use risk. Flagged Apify and Hugging Face scripts talk to those vendors with user-supplied tokens rather than a hidden collector.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
The user runs npx agentic-awesome-skills or the published bin, not npm install.
Impact
Agent instruction sets can be expanded or replaced, including offensive catalog skills if the user accepts a full or selected install.
Mechanism
Explicit CLI copies skill files into AI-agent skill directories
AI rationale
This package does not run on npm install, but the published CLI is built to copy a large mixed-risk skill catalog into AI-agent skill directories, and a binary PDF plus incomplete static coverage leave residual dual-use risk. Flagged Apify and Hugging Face scripts talk to those vendors with user-supplied tokens rather than a hidden collector.
Evidence
package.jsontools/bin/install.jsskills/theme-factory/SKILL.md~/.agents/skills~/.claude/skills~/.cursor/skills
Network endpoints2
github.com/sickn33/agentic-awesome-skills.gitregistry.npmjs.org
Decision evidence
public snapshotAI called this Suspicious at 72.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- The published bin maps to an installer that copies skills into AI-agent home directories and defaults to ~/.agents/skills when no target flag is given.
- If the user targets Cursor, Claude, or similar without a selector, the installer still proceeds with the complete catalog after warning that offensive skills may be included.
- skills/theme-factory/theme-showcase.pdf is a non-UTF8 binary that cannot be inspected as source, so its contents stay unverified.
- The published files list ships the entire skills tree, which is too large for complete static review and includes dual-use pentest documentation.
Evidence against
- package.json has no preinstall, install, or postinstall scripts.
- The installer, aas, and aas-mcp bins run only when the user launches them, not on npm install.
- The default Antigravity path refuses a full catalog write unless --all or a skill selector is supplied.
- The Apify helper sends APIFY_TOKEN only to api.apify.com after the user runs that skill script.
- The Hugging Face sample sends HF_TOKEN only to huggingface.co/api/models.
- fedora-hyprland tests/test-scripts.sh is a dry-run syntax and backup harness under an isolated HOME.
Source & flagged code
11 flagged · loading sourceskills/developer-signup-flow/SKILL.mdView file
165patternName = stripe_test_secret
severity = high
line = 165
High
High Secret
Package contains a high-severity secret pattern.
skills/developer-signup-flow/SKILL.mdView on unpkg · L165165patternName = stripe_test_secret
severity = high
line = 165
High
Secret Pattern
Stripe test secret key in skills/developer-signup-flow/SKILL.md
skills/developer-signup-flow/SKILL.mdView on unpkg · L165node_modules/ajv/dist/compile/jtd/serialize.jsView file
49sourceCode = `${gen.scopeRefs(names_1.default.scope)}return ${serializeFuncCode}`;
L50: const makeSerialize = new Function(`${names_1.default.scope}`, sourceCode);
L51: const serialize = makeSerialize(this.scope.get());
Low
Findings
1 Critical6 High6 Medium5 Low
CriticalHardcoded Runtime Data Exfiltrationskills/apify-content-analytics/reference/scripts/run_actor.js
HighHigh Secretskills/developer-signup-flow/SKILL.md
HighCredential Exfiltrationskills/huggingface-tool-builder/references/baseline_hf_api.tsx
HighShips High Entropy Blobskills/theme-factory/theme-showcase.pdf
HighPayload In Excluded Dirskills/fedora-hyprland-installer/tests/test-scripts.sh
HighSemantic Analysis Limitednode_modules/ajv/dist/refs/data.json
HighSecret Patternskills/developer-signup-flow/SKILL.md
MediumDynamic Requiretools/bin/aas.js
MediumNetwork
MediumEnvironment Vars
MediumShips Build Helperskills/gemini-deep-research/scripts/research.py
Affected versions and remediation
This report applies to agentic-awesome-skills@17.3.0.
See all versions of agentic-awesome-skills for other recorded verdicts.
- Review the evidence and your use of agentic-awesome-skills@17.3.0 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.