gate-review
Fail
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: HIGHDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: Path traversal and absolute path injection vulnerability in
scripts/check-citations.py. The script parses file paths (referred to ascorpus-path) from the## Decision-Token Legendtable in the document being audited. It resolves these paths using thepathliboperationroot / path_part. Sincepath_partis taken directly from the untrusted document content, an attacker can provide an absolute path (e.g.,/etc/passwd) or a traversal path (e.g.,../../.ssh/id_rsa) whichpathlibwill resolve to a location outside the workspace root. The script then reads the file's content to extract titles or JSON data, potentially exposing sensitive information.\n - Evidence: In
scripts/check-citations.py, theresolve_pointerfunction accepts a pointer string, extracts thepath_part, and constructsfpath = root / path_part. The file is then read viafpath.read_text()in theextract_md_titlefunction.\n- [COMMAND_EXECUTION]: The skill invokes several local scripts to perform audit tasks, includingscripts/check-citations.py,scripts/check-inheritance.py, andscripts/check-review-header.py. These are executed via the shell using theuv runcommand. Additionally, the skill references a shared utility script${CLAUDE_PLUGIN_ROOT}/skills/_shared/now.mjsfor timestamp generation.\n- [PROMPT_INJECTION]: The skill is vulnerable to Indirect Prompt Injection (Category 8).\n - Ingestion points: The skill loads and parses a
*-gate.mdfile from the workspace root, as specified in theInputsection ofSKILL.md.\n - Boundary markers: None. The skill lacks delimiters or explicit warnings to the agent to treat the document content as data rather than instructions.\n
- Capability inventory: The skill can execute shell commands for its Python and Node.js scripts, write finding reports back to the workspace files, and access the host file system via the vulnerability in its citation script.\n
- Sanitization: None. The skill does not filter or sanitize the contents of the audited document before processing it or passing its contents to shell scripts.
Recommendations
- AI detected serious security threats
Audit Metadata