go-no-go
Pass
Audited by Gen Agent Trust Hub on Jul 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The installation instructions in README.md specify using 'npx skills@latest' to add the skill, which executes code downloaded from the NPM registry at runtime.\n- [COMMAND_EXECUTION]: The skill logic relies on the agent's ability to execute system commands such as 'git log --grep' for memory searches and the 'gh' CLI for managing GitHub Gists.\n- [DATA_EXFILTRATION]: The skill features a 'public commitment' system, enabled by default, which automatically uploads project artifacts and conversation transcripts to external platforms including GitHub Gists, GitHub repositories, Substack, and Twitter.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection during 'Step 0: Pre-flight Memory Check'. The agent is instructed to read and process content from local journal files or external knowledge vaults (Obsidian, Logseq, Atomos) without the use of boundary markers or sanitization to protect against malicious instructions embedded in those files.\n- [EXTERNAL_DOWNLOADS]: The skill's documentation and configuration templates recommend and provide links to third-party services and software, such as Atomos (orangebot.ai), Obsidian, and Logseq.
Audit Metadata