ignore
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill scans local project files like
package.json,Cargo.toml, andrequirements.txtto identify the project's tech stack. This data ingestion represents a surface for indirect prompt injection where an attacker could theoretically attempt to influence the agent's behavior through crafted file names or metadata, although the impact is restricted to.gitignoreconfiguration. - Ingestion points: Project root files and configuration markers (Step 1).
- Boundary markers: No explicit delimiter or warning system for the ingested project data is defined.
- Capability inventory: The skill is limited to file system reads (Step 1, Step 4) and file system writes (Step 4).
- Sanitization: There is no mention of sanitizing or validating the contents of the detected project files before using them as signals for stack identification.
- [SAFE]: The skill implements protective measures by explicitly instructing the agent to ignore sensitive environment variables and providing a backup mechanism (
.gitignore.bak) to ensure the original configuration remains recoverable after modifications.
Audit Metadata