session-analyst
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local shell script located at
~/.claude/evolution/bin/evolve-collect.shto update candidate learnings based on fresh session records. - [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from interaction history, creating a surface where malicious instructions in previous messages could influence the generated report.
- Ingestion points: Reads event records from
~/.claude/evolution/records/*.jsonlfiles. - Boundary markers: None identified; instructions do not specify delimiters to separate interaction data from the analysis instructions.
- Capability inventory: Ability to write files to
~/.claude/evolution/reports/and execute local bash scripts. - Sanitization: No explicit validation or filtering of the content from the records is mentioned before it is summarized or used to trigger candidate pattern identification.
Audit Metadata