add-issue
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill reads untrusted data from local application logs and repository source code to generate issue descriptions.
- Ingestion points: Ingests data from terminal output, development server logs, and application files as specified in SKILL.md.
- Boundary markers: Instructions require converting inaccessible facts into sanitized fixtures but do not enforce strict structural delimiters for all ingested text.
- Capability inventory: The skill can execute shell commands via Git/GitHub CLI and write to external GitHub repositories via scripts/github.mjs.
- Sanitization: Explicit instructions are provided to the agent to redact secrets and credentials from all logs and artifacts before external exposure.
- [DATA_EXFILTRATION]: The skill transmits repository evidence, logs, and screenshots to GitHub to create issues. This is the intended purpose of the skill and is constrained to the user's authenticated GitHub environment.
- [COMMAND_EXECUTION]: Local shell commands (git, gh) and internal Node.js scripts are invoked to inspect the workspace and interact with GitHub APIs. These operations are conducted using standard tools without shell interpolation risk.
- [EXTERNAL_DOWNLOADS]: The skill communicates with GitHub APIs to fetch issue metadata and check for duplicates. These network operations target a well-known service (GitHub) for legitimate development functionality.
Audit Metadata