rea-grilling
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill performs an 'Initial Artifact Triage' phase where it executes diagnostic tools such as
readelf,strings,grep, andfileon user-provided binaries or source code. The output of these tools is ingested by the agent to formulate recommendations and interview questions. This creates a surface for indirect prompt injection if an attacker-controlled artifact contains malicious instructions in metadata or string tables. - Ingestion points: Diagnostic tool output from
readelf,DIE,strings,open_binary,inspect_pe_resources,grep, andfileas described in the 'Initial Artifact Triage' section ofSKILL.md. - Boundary markers: None specified for the ingestion of tool outputs.
- Capability inventory: The skill manages interactive scoping and triggers a '5-Stage REA Pipeline' for automated code recovery.
- Sanitization: The skill includes positive security practices for sanitizing extracted credentials (API keys, tokens) into
.env.exampletemplates, but does not define sanitization or filtering for the strings and metadata extracted during the triage phase.
Audit Metadata