post-mortem
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill operates solely through markdown instructions and templates. It does not include any executable scripts (Python, Node.js, shell) or binary files, and it does not define any external dependencies.
- [SAFE]: File system operations are restricted to writing logs to a dedicated directory (~/.claude/postmortems/) for record-keeping purposes, which is consistent with the skill's stated utility as a decision-tracking tool.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by design, as it ingests decision records and execution plans generated by other components in the workflow. However, given the absence of network capabilities or high-privilege execution tools, this surface does not present a significant security risk.
- Ingestion points: Reads previously saved decision records and execution plans from the local environment.
- Boundary markers: No explicit delimiters or warnings to ignore embedded instructions are defined in the template.
- Capability inventory: Capability is limited to local file system writes in the user's home directory.
- Sanitization: No sanitization of the content from previous records is specified before being processed into the post-mortem report.
Audit Metadata