- name
SKILL.md
技能檔案保留在發布者的儲存庫中。
在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add github/awesome-copilot --skill agent-owasp-compliance -a cursor -y在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add github/awesome-copilot --skill agent-owasp-compliance -a claude-code -y在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add github/awesome-copilot --skill agent-owasp-compliance -a codex -y在終端機中貼上複製的指令。它會將這個 SKILL.md 安裝到對應的代理中。
npx skills add github/awesome-copilot --skill agent-owasp-compliance -a grok -y在 Grok Bot 聊天中傳送複製的訊息。Bot 會從 GitHub 取得這個 SKILL.md,並加入你所有 Bot 共用的技能中。
Install this skill: https://github.com/github/awesome-copilot/tree/HEAD/skills/agent-owasp-compliance來自發布者
Evaluate AI agent systems against the OWASP Agentic Security Initiative (ASI) Top 10 — the industry standard for agent security posture.
The OWASP ASI Top 10 defines the critical security risks specific to autonomous AI agents — not LLMs, not chatbots, but agents that call tools, access systems, and act on behalf of users. This skill checks whether your agent implementation addresses each risk.
Codebase → Scan for each ASI control:
ASI-01: Prompt Injection Protection
ASI-02: Tool Use Governance
ASI-03: Agency Boundaries
ASI-04: Escalation Controls
ASI-05: Trust Boundary Enforcement
ASI-06: Logging & Audit
ASI-07: Identity Management
ASI-08: Policy Integrity
ASI-09: Supply Chain Verification
ASI-10: Behavioral Monitoring
→ Generate Compliance Report (X/10 covered)
Look for input validation that runs before tool execution, not after LLM generation.
import re
from pathlib import Path
def check_asi_01(project_path: str) -> dict:
"""ASI-01: Is user input validated before reaching tool execution?"""
positive_patterns = [
"input_validation", "validate_input", "sanitize",
"classify_intent", "prompt_injection", "threat_detect",
"PolicyEvaluator", "PolicyEngine", "check_content",
]
negative_patterns = [
r"eval\(", r"exec\(", r"subprocess\.run\(.*shell=True",
r"os\.system\(",
]
# Scan Python files for signals
root = Path(project_path)
positive_matches = []
negative_matches = []
for py_file in root.rglob("*.py"):
content = py_file.read_text(errors="ignore")
for pattern in positive_patterns:
if pattern in content:
positive_matches.append(f"{py_file.name}: {pattern}")
for pattern in negative_patterns:
if re.search(pattern, content):
negative_matches.append(f"{py_file.name}: {pattern}")
positive_found = len(positive_matches) > 0
negative_found = len(negative_matches) > 0
return {
"risk": "ASI-01",
"name": "Prompt Injection",
"status": "pass" if positive_found and not negative_found else "fail",
"controls_found": positive_matches,
"vulnerabilities": negative_matches,
"recommendation": "Add input validation before tool execution, not just output filtering"
}
What passing looks like:
# GOOD: Validate before tool execution
result = policy_engine.evaluate(user_input)
if result.action == "deny":
return "Request blocked by policy"
tool_result = await execute_tool(validated_input)
What failing looks like:
# BAD: User input goes directly to tool
tool_result = await execute_tool(user_input) # No validation
Verify tools have allowlists, argument validation, and no unrestricted execution.
What to search for:
subprocess.run(shell=True) with user-controlled inputeval() or exec() on agent-generated code without sandboxPassing example:
ALLOWED_TOOLS = {"search", "read_file", "create_ticket"}
def execute_tool(name: str, args: dict):
if name not in ALLOWED_TOOLS:
raise PermissionError(f"Tool '{name}' not in allowlist")
# validate args...
return tools[name](**validated_args)
Verify agent capabilities are bounded — not open-ended.
What to search for:
Failing: Agent has access to all tools by default. Passing: Agent capabilities defined as a fixed allowlist, unknown tools denied.
Verify agents cannot promote their own privileges.
What to search for:
Failing: Agent can modify its own configuration or permissions. Passing: Privilege changes require out-of-band approval (e.g., Ring 0 requires SRE attestation).
In multi-agent systems, verify that agents verify each other's identity before accepting instructions.
What to search for:
Passing example:
def accept_task(sender_id: str, task: dict):
trust = trust_registry.get_trust(sender_id)
if not trust.meets_threshold(0.7):
raise PermissionError(f"Agent {sender_id} trust too low: {trust.current()}")
if not verify_signature(task, sender_id):
raise SecurityError("Task signature verification failed")
return process_task(task)
Verify all agent actions produce structured, tamper-evident audit entries.
What to search for:
Failing: Agent actions logged via print() or not logged at all.
Passing: Structured JSONL audit trail with chain hashes, exported to secure storage.
Verify agents have cryptographic identity, not just string names.
Failing indicators:
agent_name = "my-agent" (string only)Passing indicators:
did:web:, did:key:)Verify policy enforcement is deterministic — not LLM-based.
What to search for:
Failing: Agent decides its own permissions via prompt ("Am I allowed to...?"). Passing: PolicyEvaluator.evaluate() returns allow/deny in <0.1ms, no LLM involved.
Verify agent plugins and tools have integrity verification.
What to search for:
INTEGRITY.json or manifest files with SHA-256 hashes@latest, >= without upper bound)Verify the system can detect and respond to agent behavioral drift.
What to search for:
Failing: No mechanism to stop a misbehaving agent automatically. Passing: Circuit breaker trips after N failures, trust decays without activity, kill switch available.
# OWASP ASI Compliance Report
Generated: 2026-04-01
Project: my-agent-system
## Summary: 7/10 Controls Covered
| Risk | Status | Finding |
|------|--------|---------|
| ASI-01 Prompt Injection | PASS | PolicyEngine validates input before tool calls |
| ASI-02 Insecure Tool Use | PASS | Tool allowlist enforced in governance.py |
| ASI-03 Excessive Agency | PASS | Execution rings limit capabilities |
| ASI-04 Unauthorized Escalation | PASS | Ring promotion requires attestation |
| ASI-05 Trust Boundary | FAIL | No identity verification between agents |
| ASI-06 Insufficient Logging | PASS | AuditChain with SHA-256 chain hashes |
| ASI-07 Insecure Identity | FAIL | Agents use string names, no crypto identity |
| ASI-08 Policy Bypass | PASS | Deterministic PolicyEvaluator, no LLM in path |
| ASI-09 Supply Chain | FAIL | No integrity manifests or plugin signing |
| ASI-10 Behavioral Anomaly | PASS | Circuit breakers and trust decay active |
## Critical Gaps
- ASI-05: Add agent identity verification using DIDs or signed tokens
- ASI-07: Replace string agent names with cryptographic identity
- ASI-09: Generate INTEGRITY.json manifests for all plugins
## Recommendation
Install agent-governance-toolkit for reference implementations of all 10 controls:
pip install agent-governance-toolkit
Use these to rapidly assess an agent system:
If you answer "no" to any of these, that's a gap to address.
|
---name: agent-owasp-compliancedescription: |---# skills/agent-owasp-compliance/SKILL.md常見問題
從發布者下載 skills/agent-owasp-compliance/SKILL.md,或複製路徑後開啟儲存庫。PlugMCP 只收錄這項技能,不代管檔案。
發布者將它存放在 https://github.com/github/awesome-copilot。
Claude、Cursor、Codex 和 Grok Build 可以讀取你從發布者儲存庫加入的 SKILL.md。在聊天中傳送技能的 GitHub 資料夾連結,Grok Bot 就會安裝它。