context-shards
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to run a Python script located at
../../scripts/context_shards.pyto perform deterministic state changes and validation checks. - [INDIRECT_PROMPT_INJECTION]: The skill incorporates an ingestion surface for untrusted data which could potentially contain malicious instructions intended to influence the agent's behavior.
- Ingestion points: The instructions require the agent to read project context from
../../references/CONTEXT_SHARDS.md. - Boundary markers: No delimiters or 'ignore' instructions are present to prevent the agent from obeying instructions embedded within the ingested files.
- Capability inventory: The agent is authorized to execute scripts and write results to the file system, providing a functional path for injected instructions.
- Sanitization: There is no evidence of validation, escaping, or filtering of the data retrieved from the reference documents.
Audit Metadata