night-market-completion-integrity-campaign
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous instructions to execute local diagnostic commands, including running
pytestsuites for internal plugins and probing hook scripts with specific JSON payloads. These operations are strictly limited to auditing the security and correctness of the local repository state. - [DYNAMIC_EXECUTION]: Python scripts are generated via shell heredocs to process the
.egregore/manifest.jsonfile. This dynamic generation is used for safe data processing and status counting within the campaign's monitoring phases. - [INDIRECT_PROMPT_INJECTION]: The skill proactively identifies attack surfaces where conversational outputs could be used to bypass integrity checks (e.g., 'talking past' a judge). It mandates documenting these vulnerabilities and implementing deterministic code-side enforcement to prevent agents from manipulating their own status manifests, explicitly addressing supply chain and autonomy risks.
Audit Metadata