claw-wiki Security Report — Low Risk | ClawSafe

20 /100

claw-wiki

Reference and refresh a local OpenClaw documentation snapshot for product usage, deployment, configuration, CLI commands, channels, gateway behavior, tools, troubleshooting, and docs-version verification.

A documentation lookup and sync skill with well-defined safety boundaries; legitimate use of subprocess for git clone is properly sandboxed to the upstream OpenClaw repo, with no credential access or exfiltration paths.

Skill Nameclaw-wiki

Duration53.1s

Enginepi

✓

Safe to install

Skill is safe to use. The flagged CRITICAL IOCs in the pre-scan are documentation examples within openclaw_docs/ content, not executable code paths. Consider adding explicit subprocess/git-clone disclosure to SKILL.md for full transparency.

Findings 3 items

Severity	Finding	Location
Low	Subprocess/git-clone not explicitly declared in allowed behavior Doc Mismatch The SKILL.md safety contract lists 'Allowed behavior' but does not mention subprocess usage or git-clone for syncing docs. The subprocess call in sync_docs.py is restricted to the pinned upstream OpenClaw repository, but transparency would be improved by explicit disclosure. `Allowed behavior: Read the bundled local docs snapshot, Search the local snapshot, When the user explicitly asks for a refresh, sync docs from the upstream OpenClaw documentation source` → Add 'git clone' to the allowed behavior section or clarify that the sync scripts use subprocess for local git operations from the pinned upstream only.	`SKILL.md:1`
Low	Dangerous shell examples embedded in documentation content Sensitive Access The openclaw_docs/ content includes shell command examples (e.g., curl\|bash install scripts, rm -rf commands) that are flagged as CRITICAL IOCs in the pre-scan. These are legitimate documentation examples about OpenClaw installation and are NOT executed by the skill's scripts. SKILL.md explicitly disclaims treating docs examples as real credentials or executing install scripts. `curl -fsSL --proto '=https' --tlsv1.2 https://openclaw.ai/install-cli.sh \| bash` → No action needed; these are documentation examples. The skill's hard limits correctly prohibit executing downloaded install scripts.	`openclaw_docs/install/installer.md:35`
Info	No dependency pinning in scripts Supply Chain The Python scripts use only stdlib modules (subprocess, shutil, json, pathlib, argparse, re, hashlib) with no external package dependencies declared in requirements.txt or similar. This eliminates supply-chain risk from third-party packages. `from __future__ import annotations from collections import defaultdict from datetime import datetime, timezone` → No action needed; stdlib-only dependencies are safe.	`scripts/build_knowledge.py:1`

Resource	Declared	Inferred	Status	Evidence
Filesystem	`NONE`	`WRITE`	✓ Aligned	sync_docs.py:89 — writes to skill_root/openclaw_docs/ via shutil, confined by en…
Shell	`NONE`	`WRITE`	✓ Aligned	sync_docs.py:69 — git clone via subprocess from pinned upstream URL only, not me…
Network	`READ`	`READ`	✓ Aligned	sync_docs.py:69 — network access is read-only (git clone), limited to upstream O…
Environment	`NONE`	`NONE`	—	No os.environ iteration or environment variable access found in any script
Skill Invoke	`NONE`	`NONE`	—	No cross-skill invocation observed
Clipboard	`NONE`	`NONE`	—	No clipboard access in any script
Browser	`NONE`	`NONE`	—	No browser tool usage
Database	`NONE`	`NONE`	—	No database access

11 Critical 21 High 396 findings

💀