cost-analysis
Warn
Audited by Gen Agent Trust Hub on Mar 23, 2026
Risk Level: MEDIUMDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill accesses highly sensitive files including .env, local.settings.json, terraform.tfvars, and various cloud provider configuration manifests (AWS, GCP, Azure). These files are primary locations for hardcoded credentials, API keys, and sensitive architectural metadata.
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it is instructed to 'Read the actual codebase' and 'Discover every cost-generating service' without implementing boundary markers or sanitization. Malicious instructions placed in the codebase or configuration files could be followed by the agent during the analysis process.
- [PROMPT_INJECTION]: Mandatory Evidence Chain: * Ingestion points: Entire project codebase, cloud service configuration files (firebase.json, serverless.yml, etc.), and dependency manifests. * Boundary markers: Absent; no instructions provided to ignore or delimit embedded prompts in data. * Capability inventory: Recursive file read access, file creation and writing (docs/cost-analysis.md), and potential web search for pricing verification. * Sanitization: Absent; the skill directly parses and uses data found in the repository.
Audit Metadata