gitwhy-context-saving
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFEDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes historical context (reasoning, stories, and decisions) which could contain adversarial instructions if the source context or repository was compromised.
- Ingestion points: Engineering context is retrieved via
gitwhy_get,gitwhy_search, andgitwhy_listfrom local files in.git/gitwhy/contexts/or the cloud API. - Boundary markers: While XML tags are used for initial context input, retrieved content is presented as standard Markdown without explicit delimiters or instructions to ignore embedded commands.
- Capability inventory: The skill has the capability to execute shell commands (
git why), perform file system operations, and communicate with an external API. - Sanitization: No explicit content sanitization or instruction-filtering mechanisms are documented for the context retrieval and rendering process.
- [COMMAND_EXECUTION]: Interacts with the local system by executing the
git whyCLI tool to perform context logging, searching, and configuration. - [CREDENTIALS_UNSAFE]: Accesses an API key stored in the user's home directory at
~/.gitwhy/credentialsfor authenticating cloud operations. This represents standard credential management for the vendor's tool. - [DATA_EXFILTRATION]: Provides functionality to synchronize local context data and engineering reasoning to the vendor's cloud service at
app.gitwhy.devand to post updates to GitHub Pull Requests.
Audit Metadata