context-save
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes a local utility script
python tools/allocate-id.pyto ensure atomic ID generation across different namespaces (Decisions, Bets, Assumptions, Learnings). This is a localized operation within the skill's workspace and does not involve external command sources. - [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it ingests and processes untrusted data from the conversation history or user-provided files to extract metadata for the registry.
- Ingestion points: Ingests recent conversation text and external files referenced via the
@pathsyntax (specifically inSKILL.md). - Boundary markers: None explicitly defined in the extraction logic to separate data from instructions.
- Capability inventory: The skill possesses file-write capabilities to the
context/directory and executes a local Python script via subprocess. - Sanitization: The instructions do not specify sanitization or escaping of the content before it is processed or written to index files.
Audit Metadata