cs-doc-tutorial
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes a local Python utility script
.codestable/tools/search-yaml.pyto search and filter documentation metadata. This script is part of the vendor's local toolset and does not involve remote code execution. - [PROMPT_INJECTION]: Instructions include directives to 'ignore external AI entry replacements' and prioritize specific project conventions (
.codestable/reference/execution-conventions.md). These are operational scoping instructions rather than attempts to bypass security guardrails. - [DATA_EXFILTRATION]: The skill reads from local project files like
.codestable/attention.mdand writes documentation todocs/dev/anddocs/user/. There is no evidence of network activity, exfiltration of sensitive system credentials, or access to sensitive paths outside the project directory. - [SAFE]: The skill's primary purpose is generating structured documentation based on local inputs. The use of local scripts and specific project paths is consistent with its stated purpose as a developer tool.
Audit Metadata