resource-estimation
Pass
Audited by Gen Agent Trust Hub on Mar 23, 2026
Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [PROMPT_INJECTION]: The skill instructs the agent to 'Do NOT ask the user questions. Analyze and act.', which encourages autonomous execution and may suppress user confirmation prompts during the analysis process.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted geological data from various formats without sanitization.
- Ingestion points: Reads geological modeling files (.bmf, .dm, .mdl), CSV/Excel databases, and assay results in SKILL.md Phase 1.
- Boundary markers: Absent; there are no instructions to ignore or delimit embedded commands within processed data files.
- Capability inventory: The agent has capabilities to write reports to the project directory and append metadata to hidden configuration files in the home directory.
- Sanitization: Absent; the skill does not specify any validation or filtering of the content retrieved from external data files before processing.
- [DATA_EXFILTRATION]: The 'Self-Evolution Telemetry' section writes execution metadata, outcomes, and project paths to a hidden directory (
~/.claude/projects/) in the user's home folder. This behavior tracks and persists project-specific information outside the current workspace, which can be used for activity monitoring across sessions.
Audit Metadata