中文 EN

Scan Report

Scan New Files

✓ Trusted — Risk Score 5/100
Last scan:2 days ago Rescan
5 /100
figma-agent
Unified Figma skill for OpenClaw. Reads design context via Figma Remote MCP and routes write/edit/create operations through Claude Code ACP sessions.
Figma Agent is a well-documented, transparent skill that bootstraps Figma MCP tokens by scanning known credential stores. All behavior—network calls, filesystem reads, token refresh, and config writing—is explicitly declared in SKILL.md with zero obfuscation or hidden functionality.
Skill Namefigma-agent
Duration62.1s
Enginepi
✓
Safe to install
Approve for use. No security concerns identified. The credential-scanning behavior is declared, scoped, and serves a legitimate bootstrap purpose.

Findings 4 items

Severity Finding Location
Info
Credential file scanning for token bootstrap
token-scanner.mjs scans three credential stores (~/.claude/.credentials.json, ~/.codex/auth.json, ~/.codeium/windsurf/mcp_config.json) to find Figma MCP tokens. This is legitimate token bootstrapping fully documented in SKILL.md Token management section. Only Figma-specific tokens are extracted and written to OpenClaw config. No exfiltration observed.
const claudePath = join(HOME, '.claude', '.credentials.json');
→ No action needed. Behavior is declared and scoped to Figma tokens only.
 scripts/token-scanner.mjs:23
Info
OAuth token refresh via Figma API
bootstrap-token.mjs makes HTTP POST to api.figma.com/v1/oauth/token for token refresh. This is standard OAuth behavior and the only network call to external services (alongside mcp.figma.com/mcp). Both are declared.
const metaRes = await fetch('https://api.figma.com/v1/oauth/token');
→ No action needed. Standard OAuth token refresh, declared in SKILL.md.
 scripts/bootstrap-token.mjs:32
Info
No dependencies
package.json has zero dependencies (dependencies: {}). All functionality uses built-in Node.js APIs (fs, os, path, fetch). No supply chain risk.
"dependencies": {}
→ Excellent dependency hygiene. Maintain this standard.
 package.json:1
Info
Config write only to OpenClaw locations
writeToOpenClaw() in bootstrap-token.mjs writes the token only to known OpenClaw config paths (~/.openclaw/openclaw.json or ~/.config/openclaw/openclaw.json). No arbitrary file writes.
config.mcp.servers.figma.headers = { Authorization: `Bearer ${token}` };
→ No action needed. Write scope is constrained to OpenClaw config.
 scripts/bootstrap-token.mjs:75
ResourceDeclaredInferredStatusEvidence
Filesystem READ READ ✓ Aligned SKILL.md: requires env FIGMA_MCP_TOKEN; token-scanner.mjs:23-89 reads ~/.claude/…
Network READ READ ✓ Aligned SKILL.md: talks to mcp.figma.com/mcp via Remote MCP; bootstrap-token.mjs:30-52 r…
Shell NONE NONE — No shell invocation found. All scripts are pure Node.js ESM modules with no subp…
Environment READ READ ✓ Aligned SKILL.md requires env FIGMA_MCP_TOKEN; figma-mcp-cli.mjs reads FIGMA_MCP_TOKEN f…
Skill Invoke ADMIN ADMIN ✓ Aligned SKILL.md declares full Figma read/write capabilities; write path routes through …
Clipboard NONE NONE — No clipboard access found
Browser NONE NONE — No browser access; generate_figma_design HTML capture limitation documented in S…
Database NONE NONE — No database access found
13 findings
🔗
Medium External URL 外部 URL
https://keepachangelog.com/
CHANGELOG.md:5
🔗
Medium External URL 外部 URL
https://semver.org/
CHANGELOG.md:5
🔗
Medium External URL 外部 URL
https://mcp.figma.com/mcp
README.md:5
🔗
Medium External URL 外部 URL
https://img.shields.io/badge/License-MIT-green.svg
README.md:9
🔗
Medium External URL 外部 URL
https://img.shields.io/badge/version-v0.1.0-blue.svg
README.md:10
🔗
Medium External URL 外部 URL
https://clawhub.ai
README.md:11
🔗
Medium External URL 外部 URL
https://img.shields.io/badge/ClawHub-skill-purple.svg
README.md:11
🔗
Medium External URL 外部 URL
https://nodejs.org
README.md:57
🔗
Medium External URL 外部 URL
https://help.figma.com/hc/en-us/articles/32132100833559
README.md:58
🔗
Medium External URL 外部 URL
https://developers.figma.com/docs/figma-mcp-server/write-to-canvas/
README.md:168
🔗
Medium External URL 外部 URL
https://api.figma.com/v1/oauth/token
scripts/bootstrap-token.mjs:32
🔗
Medium External URL 外部 URL
https://figma.com/mcp
scripts/figma-mcp.mjs:179
📧
Info Email 邮箱地址
[email protected]
package.json:6

File Tree

9 files · 34.4 KB · 946 lines
Markdown 4f · 477L JavaScript 4f · 451L JSON 1f · 18L
├─ ▾ 📁 references
│ └─ 📝 figma-api.md Markdown 35L · 1.2 KB
├─ ▾ 📁 scripts
│ ├─ 🔑 bootstrap-token.mjs ⚠ JavaScript 137L · 4.4 KB
│ ├─ 📜 figma-mcp-cli.mjs JavaScript 28L · 925 B
│ ├─ 📜 figma-mcp.mjs JavaScript 196L · 6.9 KB
│ └─ 🔑 token-scanner.mjs ⚠ JavaScript 90L · 2.8 KB
├─ 📝 CHANGELOG.md Markdown 15L · 1.1 KB
├─ 📋 package.json JSON 18L · 396 B
├─ 📝 README.md Markdown 203L · 7.5 KB
└─ 📝 SKILL.md Markdown 224L · 9.2 KB

Dependencies 1 items

PackageVersionSourceKnown VulnsNotes
(none) N/A npm No No dependencies declared in package.json — uses only built-in Node.js modules (fs, os, path, fetch)

Security Positives

✓ Zero dependencies — no supply chain risk
✓ All filesystem reads are explicitly declared in SKILL.md under Token management
✓ All network calls go to known Figma API endpoints (api.figma.com, mcp.figma.com)
✓ No obfuscation, base64, eval, or dynamic code execution
✓ No shell execution, subprocess, or remote script execution
✓ Token scanning is scope-limited to Figma tokens only (key includes 'figma' check)
✓ writeWithCheckpoint() safety pattern documented for write operations
✓ DRY_RUN flag provides safe preview before writes
✓ Clean separation: token-scanner.mjs handles file I/O, bootstrap-token.mjs handles network
✓ No hidden instructions in HTML comments or string literals
✓ AllowedTools mapping: Read→filesystem:READ, WebFetch→network:READ — matches inferred usage