ClawAudit verdict
pdf-learning-workflow
Accesses credentials AND encodes data
Explicitly declares the GLM-OCR API key is only sent to api.bigmodel.cn as an Authorization header; credential_access is limited to reading the key file, transparently documented.
Automated static analysis — not a human review. ClawAudit flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.
What it does
These are capability combinations: each listed behavior occurs in the skill, but ClawAudit detects co-occurrence — it does not verify that one flows into another. Read the code to confirm a live chain.
Accesses credentials AND encodes data — may obfuscate stolen credentials
LLM02 · ASI03 · ASI04
Permission integrity
file_read
credential_access
Findings (3)
Pipe to python — executes piped content as Python code
README.md · prose · downgraded · | Python
<script> tag in markdown — potential code injection
scripts/md2html.py · prose · downgraded · <script
pip3 install — installs Python packages at runtime
meta.json · prose · downgraded · pip3 install
Permissions & capabilities
No declared permissions — minimal attack surface.
file_readcredential_accessdata_encoding Thanks — recorded.