update-focus
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from git logs and session log files to generate summaries, which creates a potential surface for indirect prompt injection if those logs contain malicious instructions.
- Ingestion points: .context/current-focus.md, session logs in the log/ directory, and git log output.
- Boundary markers: No explicit delimiters or ignore-instructions are used for the ingested content.
- Capability inventory: The skill utilizes the Edit tool to perform file system writes.
- Sanitization: No automated sanitization is performed; however, the skill mandates a human-in-the-loop confirmation step (Step 4) where the draft summary and proposed changes must be reviewed and approved by the user before being applied.
- [COMMAND_EXECUTION]: The skill uses restricted shell commands to retrieve project metadata.
- Evidence: allowed-tools includes Bash(git log*), Bash(date*), and Bash(ls*).
- Context: These commands are used as intended to provide the agent with work history and date information necessary for accurate document rotation and updating.
Audit Metadata