Security scanner for AI agent skills. Detect vulnerabilities, malicious patterns, security risks, prompt injection, data exfiltration, and supply-chain risks in Claude Code, Codex, and MCP skills before you install them.
- capability exposure inferred + 35
- tool safety inferred + 17
- trust mitigators mixed − 3
inferred mixed
The A–E grade is our heuristic synthesis — a "review this" prompt, not a verdict. Each factor is tagged by what backs it: attested (a verifiable record), reported (a third party's claim), or inferred (our own heuristic, e.g. permissions). See methodology.
grade last moved 1w ago · see ecosystem CVEs →
- D · 69 → C · 49
- D · 61 → D · 69
No known CVEs for this server.
- high dangerous code
dynamic exec: __import__(), __import__ sink, eval()/exec() · obfuscated payload: base64 decode → exec sink
- medium toxic flow (lethal trifecta)
lethal trifecta reachable across this server's tool + source surface: private-data access + untrusted-content ingestion + network exfil (a leg is proven only in the analyzed source)
analyzed commit 548e5e0 · analyzer v33 · 1h ago
skills & prompt files 27
- agent-rules NVIDIA-SkillSpector-548e5e0/.opencode/commands/skillspector.md
- skill NVIDIA-SkillSpector-548e5e0/skills/skill-inspector/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/as3_self_reference/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/malicious_skill/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/mcp_clean_skill/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/mcp_mismatched_skill/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/mcp_overprivileged_skill/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/mcp_poisoned_tool/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/mcp_underdeclared_skill/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/pe3_bare_keyring/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/safe_skill/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sdi/sdi1_mismatch/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sdi/sdi2_inappropriate/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sdi/sdi3_scope_creep/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sdi/sdi4_divergence/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sdi/sdi_clean/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sqp/sqp1_clean/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sqp/sqp1_vague_triggers/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sqp/sqp2_clean/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sqp/sqp2_missing_warnings/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sqp/sqp3_clean/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/sqp/sqp3_locale_forcing/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/ssd/ssd1_semantic_injection/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/ssd/ssd2_novel_phrasing/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/ssd/ssd3_nl_exfiltration/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/ssd/ssd4_narrative_deception/SKILL.md
- skill NVIDIA-SkillSpector-548e5e0/tests/fixtures/ssd/ssd_clean/SKILL.md
danger signals5
- dynamic code execution __import__() NVIDIA-SkillSpector-548e5e0/contrib/batch_scan/tests/tests-pro/mutation_max.py :35
mod = __import__(module, fromlist=[""]) - dynamic code execution __import__ sink NVIDIA-SkillSpector-548e5e0/contrib/batch_scan/tests/tests-pro/test_runner_patches.py :303
__import__("os").getenv("SKIP_SLOW_TESTS"), - dynamic code execution eval()/exec() NVIDIA-SkillSpector-548e5e0/tests/unit/test_compare_scan_accuracy.py :1096
exec(compare_scan_accuracy._RUNTIME_IDENTITY_PROBE, {}) - suspicious endpoint 169.254.169.254 (cloud metadata) NVIDIA-SkillSpector-548e5e0/tests/nodes/analyzers/test_static_patterns.py :1371
'requests.get("http://169.254.169.254/latest/meta-data/iam/security-credentials/")\n' - suspicious endpoint 169.254.169.254 (cloud metadata) NVIDIA-SkillSpector-548e5e0/tests/unit/test_input_handler_ssrf.py :80
handler._clone_git("http://169.254.169.254/latest/meta-data/")
[](https://mcpobservatory.com/servers/github:NVIDIA/SkillSpector/security) Heuristic, inferred signals — false positives (legitimately powerful tools, forks, language ports) are expected. Treat each as "review this", not a verdict. See the ecosystem-wide picture on the security hub, or the fleet security of NVIDIA.