ditto
Pass
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses shell commands to run a local Python runtime. It instructs the agent to execute commands such as
"$PYTHON3" "$DITTO_PY" plugin preflightand"$PYTHON3" "$DITTO_PY" plugin prepare. The paths for these variables are resolved by asking the user for a trusted local installation path. - [PROMPT_INJECTION]: The skill is subject to indirect prompt injection because it processes untrusted data from local coding-agent session logs.
- Ingestion points: Local coding-agent session logs from Claude Code, Codex, Copilot CLI, or OpenCode (found in SKILL.md).
- Boundary markers: The skill explicitly instructs the agent to ignore
AGENTS.md, memory files, and rules files to prevent synthesis from instructions rather than user behavior. - Capability inventory: The skill utilizes subprocess calls to the local Python runtime for preflight, preparation, mining, and status reporting.
- Sanitization: The skill notes that redaction of session content is best-effort and advises the user to inspect output before sharing.
Audit Metadata