portfolio-artifact
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes several standard command-line utilities (git, jq, grep, wc, date) to extract metrics and generate markdown reports. These commands are localized to the user's development environment and are consistent with the skill's stated purpose of session metric tracking.
- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests data from git logs and conventional commit messages to populate its report templates.
- Ingestion points: SKILL.md and reference/metrics-guide.md utilize git log to retrieve commit messages.
- Boundary markers: The skill does not implement specific delimiters or warnings to ignore instructions that might be embedded in commit messages.
- Capability inventory: The skill performs file writes to ~/.claude/portfolio/ and executes local CLI tools (git, jq), but does not pass untrusted data into an interpreter like eval() or exec().
- Sanitization: There is no evidence of sanitization or escaping of the ingested commit data before it is written to JSONL or Markdown reports.
Audit Metadata