gstack-openclaw-retro
Pass
Audited by Gen Agent Trust Hub on Apr 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses shell commands including
git,find,grep, andwcto analyze the repository's commit history and file structure. These operations are read-oriented and typical for a code analysis tool. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes commit messages, which are external data that could be controlled by an attacker. A malicious commit message could attempt to manipulate the retrospective's output.
- Ingestion points: Git commit subjects and author names retrieved via
git logcommands inSKILL.md. - Boundary markers: Absent; the skill does not use specific delimiters to separate untrusted commit data from its instructions.
- Capability inventory: The skill has the ability to execute shell commands and write to the local file system (saving history to the
memory/directory) as seen inSKILL.md. - Sanitization: No sanitization or validation of the commit data is performed before it is processed by the AI.
Audit Metadata