trust-boundary-analysis
Trust boundary mapping and startup sequence audit for developer tools, CLI apps, and plugin systems. Load when the target is a developer tool, CLI, IDE extension, or any application that loads config from the current directory.
- License 许可
- Apache-2.0
- 所属仓库
- Akapi895/raptix
- 最近更新
- 2026-11-06
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- trust-boundary-analysis
- metadata
- { "subdomain": "analyst", "when_to_use": "trust boundary startup sequence developer tool cli ide extension plugin config load current directory analysis", "mitre_attack": "T1574" }
- description
- Trust boundary mapping and startup sequence audit for developer tools, CLI apps, and plugin systems. Load when the target is a developer tool, CLI, IDE extension, or any application that loads config from the current directory.
03 SKILL.md 正文
Trust Boundary Analysis
This skill targets the vulnerability class that produced 5 RCE vectors in Google Gemini CLI from a single architectural flaw: missing workspace trust. Developer tools that auto-load configuration from untrusted directories are a rich attack surface.
When to Use
Apply this skill when the target:
- Loads
.env,settings.json, or config files from the current working directory - Has a plugin/extension system that auto-discovers and loads code
- Spawns child processes with configuration-controlled commands
- Is a CLI tool, IDE extension, language server, or build tool
- Uses MCP (Model Context Protocol) servers configured via project files
Startup Sequence Audit
For any CLI tool or developer application, trace the full initialization:
1. Config file discovery
grep -rn 'readFile\|readFileSync\|fs.read\|open(' --include='*.ts' --include='*.js' \
/workspace/target/src/ | grep -i 'config\|settings\|env\|rc\|\.json'
Map the search order. Common dangerous patterns:
cwd/.tool/config.json→cwd/.env→~/.tool/config→/etc/tool/config- If the local (cwd) config is loaded BEFORE the user's global config, the attacker's repo-level config wins.
Record each config loading point:
kg_record('[{"kind": "CodeLocation", "key": "code-location::loadConfig-cwd-env",
"label": "loadConfig() reads .env from cwd",
"props": {"file": "src/config/settings.ts", "line": 42,
"trust_level": "untrusted"}}]')
2. Environment variable injection
grep -rn 'process\.env\|os\.environ\|env::var\|getenv' --include='*.ts' \
--include='*.py' --include='*.rs' /workspace/target/src/
Check: Are env vars from .env files injected into process.env? Which vars
control dangerous behavior? Look for:
*_COMMAND,*_CMD,*_EXEC→ shell execution*_PROXY→ SSRF / network interception*_PATH,*_DIR→ path traversal*_URL→ open redirect / SSRFDEBUG,NODE_ENV→ bypass security controls
3. Workspace trust check
grep -rn 'trust\|isTrusted\|workspace.*safe\|folder.*trust' --include='*.ts' \
--include='*.js' /workspace/target/src/
Look for the pattern:
// DANGEROUS: trust disabled by default
if (!settings.folderTrustEnabled) {
return { isTrusted: true }; // ← All workspaces trusted
}
If no trust check exists, every config is loaded blindly. This is the root cause.
4. Command execution from config
grep -rn 'spawn\|exec\|execSync\|child_process\|subprocess\|os\.system\|Popen' \
--include='*.ts' --include='*.js' --include='*.py' /workspace/target/src/
For each spawn/exec call, check:
- Is
shell: trueset? → shell injection via crafted values - Does the command come from config? → attacker-controlled execution
- Is there any validation/allowlist on the command?
5. Plugin/tool auto-discovery
grep -rn 'discoverTools\|loadPlugins\|autoDiscover\|mcpServers\|toolDiscovery' \
--include='*.ts' --include='*.js' /workspace/target/src/
Check: Are plugins/tools loaded from project-level config BEFORE any user confirmation? MCP servers, language server plugins, and build tool extensions are common vectors.
Trust Boundary Graph Modeling
Map the attack flow as a chain in the knowledge graph:
ENTRYPOINT (malicious .env file in cloned repo)
→ enables → ENV_INJECT (process.env poisoned)
→ enables → SHELL_EXEC (spawn with shell:true)
→ reaches → CROWN_JEWEL (arbitrary code execution)
Use low edge weights (0.2–0.4) for automatic/silent steps. Use higher weights (1.0+) for steps requiring user interaction.
Then record the chain as an AttackPath observation via kg_record;
chain planning is not on your tool surface.
Shell Quoting Vulnerabilities
When config values flow into shell commands, check for:
- Direct interpolation:
`command ${userValue}`→ injection via; id - Single-quote escape:
osascript -e '${path}'→ escape via'in path - Env var in shell context:
PROXY_CMD="curl evil.com"→ executed as-is - Argument injection:
spawn('git', ['clone', userUrl])→--upload-pack=prefix
For each, the PoC pattern is:
- Craft input that breaks out of the quoting context
- Inject a command that proves execution (
id,env, write a sentinel file) - Use
validate_findingwith the sentinel as the success pattern
Negative Control
Always test: does the same code path execute WITHOUT the malicious config? The negative control is the same operation with a clean/empty config directory. If the behavior occurs regardless, it's not a vulnerability — it's expected.
04 同仓库更多 Skills
wps-pixie-dust
WPS Pixie-Dust offline nonce attack (reaver -K / pixiewps) and fallback online PIN brute (bully) to recover the AP's WPA PSK without capturing a handshake.
wpa3-sae
WPA3-SAE transition-mode downgrade (DragonShift), SSID Confusion CVE-2023-52424, Dragonblood side-channels, and SAE captive-portal credential recovery against WPA3-Personal networks.
wpa2-psk
WPA/WPA2-PSK handshake capture via targeted deauth + PMKID (no deauth required) + offline hashcat cracking. The most common consumer/SMB encryption mode in 2026.
wpa-enterprise-eap
WPA/WPA2/WPA3-Enterprise (802.1X/EAP) rogue-RADIUS evil-twin for MSCHAPv2 capture, GTC downgrade, and PEAP relay. MSCHAPv2 capture equals a NetNTLM hash — the primary wireless on-ramp to Active Directory.
krack-fragattacks
KRACK key-reinstallation (CVE-2017-13077..13082) and FragAttacks fragmentation/aggregation flaws (CVE-2020-24586..24588, CVE-2020-26139..26147) against legacy or embedded 802.11 supplicants with poor patch cadence.
evil-twin-karma
Evil-twin rogue AP with KARMA/Mana PNL-probe response, captive-portal credential capture, and post-association MITM for PSK/open networks. Distinct from wpa-enterprise-eap which targets 802.1X.